chore(release): v0.5.0 — image provenance (identify) + SD/SDXL/FLUX + EXIF/XMP detection

New since v0.4.1: - identify command: aggregate C2PA, IPTC, SD/ComfyUI params, SynthID proxy, visible sparkle, open invisible watermark into one provenance verdict (--json, --no-visible). - Open SD/SDXL/FLUX invisible-watermark detection (imwatermark, extra: detect). - EXIF Software / XMP CreatorTool generator-tag reading (incl. AVIF/HEIF). - Stability AI + Microsoft/Bing C2PA issuers; SynthID metadata detection. - SynthID reference corpus + experimental pixel-carrier probe. - Fix: __version__ was stuck at 0.3.4 (banner mismatch), now synced. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
docs: correct SynthID spectral-carrier understanding
2026-10-14 21:40:47 +02:00 · 2026-05-25 08:58:45 -07:00 · 2026-05-25 08:57:35 -07:00 · 2026-05-24 18:38:56 -07:00 · 2026-05-24 17:56:39 -07:00 · 2026-05-24 17:12:42 -07:00
26 changed files with 2543 additions and 119 deletions
@@ -21,6 +21,9 @@ Thumbs.db
 # Test results
 data/results/

+# SynthID corpus images (manifest.csv + README.md stay tracked)
+data/synthid_corpus/images/
+
 # Reference materials
 _refs/

@@ -30,3 +33,4 @@ yolov8n.pt

 # Claude Code local settings
 .claude/settings.local.json
+data/synthid_corpus/refs/
@@ -5,9 +5,18 @@ You are a **principal Python engineer** maintaining a CLI tool and library for r
 ## How to run

 - `uv run remove-ai-watermarks all <image.png> -o <output.png>`
+- `uv run remove-ai-watermarks identify <image>` — provenance verdict (platform + watermark inventory + confidence); `--json` for machine output, `--no-visible` to skip the cv2 sparkle detector
 - `uv run remove-ai-watermarks metadata <image.png> --check` — inspect AI metadata (C2PA, EXIF, PNG chunks)
 - `uv run remove-ai-watermarks metadata <image.png> --remove -o <out.png>` — strip all AI metadata

+## Test and lint
+
+- `bash maintain.sh` — uv-outdated, uv-secure, ruff check/fix, ruff format, pyright, pytest -n auto
+- `maintain.sh` does not currently finish green (pre-existing, not per-change): `uv-secure` aborts on a fixable transitive `idna` vuln, and strict pyright carries debt in `remove_ai_metadata` / `cli.py` (untyped piexif/PIL/click/rich). To gate a change, run `uv run ruff check`, `uv run pyright <changed files>`, `uv run pytest` directly.
+- Run `uv run` from the repo root — from another cwd it falls back to a bare env without numpy/cv2/torch.
+- Metadata/C2PA tests assert against real committed fixtures in `data/samples/` (`chatgpt-*.png` = OpenAI C2PA, `firefly-1.png` = Adobe, `not-ai-*` = clean); synthetic byte blobs cover the JPEG/ISOBMFF format paths.
+- SynthID reference corpus: `scripts/synthid_corpus.py` ingests labeled images into `data/synthid_corpus/` (`manifest.csv` tracked, `images/` gitignored); see its README for the collection protocol and verification oracles.
+
 ## Configuration

 - GPU/ML modules (invisible_engine, ctrlregen, watermark_remover) are optional — guard imports with `is_available()` checks
@@ -15,14 +24,29 @@ You are a **principal Python engineer** maintaining a CLI tool and library for r

 ## Key modules

- `noai/c2pa.py` — PNG chunk parser; use `extract_c2pa_chunk(path)` to get raw caBX payload, `has_c2pa_metadata(path)` to detect. Do not reimplement chunk parsing.
- `noai/constants.py` — PNG_SIGNATURE, C2PA_CHUNK_TYPE, C2PA_SIGNATURES constants
+- `noai/c2pa.py` — PNG chunk parser; use `extract_c2pa_chunk(path)` to get raw caBX payload, `has_c2pa_metadata(path)` to detect. Do not reimplement chunk parsing. `extract_c2pa_info(path)` sets `synthid_watermark`/`synthid_vendors` when the manifest is signed by a SynthID-using vendor.
+- `noai/constants.py` — PNG_SIGNATURE, C2PA_CHUNK_TYPE, C2PA_SIGNATURES, C2PA_ISSUERS, and `SYNTHID_C2PA_ISSUERS` (issuers that pair SynthID with C2PA: Google, OpenAI). Add a new issuer here, not inline.
+- `metadata.py` — `synthid_source(path)` returns the vendor name(s) if the C2PA manifest implies a SynthID pixel watermark, else None. Format-agnostic: PNG via the caBX parser, JPEG/WebP/AVIF/HEIF/JXL via a binary scan (C2PA marker + SynthID issuer + AI-source marker). `get_ai_metadata` surfaces the verdict, and `metadata --check` prints it as a callout. Both `get_ai_metadata` and `has_ai_metadata` guard the PIL open with `except Exception` (HEIC/unknown formats raise non-OSError) and fall through to the binary scan.
+- `identify.py` — `identify(path)` aggregates every locally-readable signal (C2PA issuer→platform, IPTC "Made with AI", embedded SD/ComfyUI params, SynthID proxy, visible Gemini sparkle) into one `ProvenanceReport`. `is_ai_generated` is True or None (never asserted False — stripped metadata is not proof of clean origin). Visible-sparkle is promoted only at confidence ≥ `_SPARKLE_THRESHOLD` (0.5; corpus-tuned to separate Gemini sparkles ≥0.56 from non-sparkle ≤0.49). The cv2 dependency lives in `gemini_engine.detect_sparkle_confidence`, not here. Add platform mappings to `_ISSUER_PLATFORM`, not inline. For non-PNG containers (JPEG/WebP/AVIF/HEIF/JXL) the caBX parser returns nothing, so issuer (`_issuers_in`) and generator (`_ai_tools_in`, reusing `C2PA_AI_TOOLS`) are recovered by binary-scanning the first MB. EXIF `Software` / `Make` / `Artist` / `ImageDescription` and XMP `CreatorTool` generator tags are read by `metadata.exif_generator` (PIL+piexif for any format PIL opens incl. AVIF, plus a container-agnostic XMP raw-byte scan that also covers HEIF/JXL), matched against `AI_GENERATOR_TOKENS` so ordinary editors (plain "Adobe Photoshop") and real-camera `Make` ("Apple"/"Canon") are not flagged. **Ideogram tags its output with EXIF `Make="Ideogram AI"`** (verified on a real download 2026-05-24) — that's why `Make` is read.
+- `gemini_engine.py` — visible Gemini-sparkle remover/detector (cv2/numpy, no GPU). `detect_sparkle_confidence(path)` is the file-level entry point used by `identify.py`.
+- `invisible_watermark.py` — `detect_invisible_watermark(path)` decodes the OPEN DWT-DCT watermarks (public decoder, no key) embedded by Stable Diffusion / SDXL / FLUX via the `imwatermark` library. Known fixed patterns (verified against upstream source) live in `_BITS_48` (SDXL 48-bit, FLUX.2 48-bit) and `_SD1_STRING` ("StableDiffusionV1", SD 1.x/2.x). Optional dep (extra `detect`); returns None when absent. **Unlike SynthID this is locally detectable**, but the watermark is fragile (does not survive JPEG re-encode/resize — verified gone after JPEG q90), so it confirms origin only on pristine files. Add new known patterns here. The file carries a top-of-module pyright pragma because imwatermark/cv2 ship no type stubs.
 - `face_protector.py` — YOLO detect + soft-blend pattern; mirror this for any "protect region during diffusion" features

+## Watermarking landscape (research 2026-05-24)
+
+Who embeds what, and whether it is locally detectable (so we know which gaps are fillable). See `identify.py` for what we read.
+- **Locally detectable (open decoder, no key/API):** Stable Diffusion / SDXL / FLUX via `imwatermark` DWT-DCT (now covered by `invisible_watermark.py`). FLUX uses the same library (`black-forest-labs/flux2` `src/flux2/watermark.py`, 48-bit `0b001010101111111010000111100111001111010100101110`); SDXL is the diffusers `WATERMARK_MESSAGE` (`0b101100111110110010010000011110111011000110011110`). Caveat: fragile to re-encoding.
+- **C2PA / IPTC (covered by the issuer/marker scan):** OpenAI, Google, Adobe Firefly, Microsoft (Designer + **Bing Image Creator** — collected 2026-05-24; Bing now runs Microsoft's own **MAI-Image** model, signs C2PA as "Microsoft", NOT OpenAI/DALL-E), and **Stability AI** (collected from Brand Studio / DreamStudio successor; signs C2PA as "Stability AI Ltd", no SynthID, no imwatermark on its current Stable Image model — issuer added to `C2PA_ISSUERS`). Still unsampled: Canva (its downloads are re-encoded design *exports* that strip C2PA, so a Canva "positive" is inconclusive — skipped), Getty, Shutterstock. Midjourney embeds NO C2PA and no invisible watermark (our `mj-*` sample carried only the IPTC tag).
+- **EXIF/XMP generator tag (caught by `exif_generator`):** **Ideogram** writes EXIF `Make="Ideogram AI"` (collected 2026-05-24 — no C2PA, no SynthID, no imwatermark; the Make tag is the only signal).
+- **No detectable signal on download (correctly reported `unknown`):** **Recraft** (PNG export is a re-encoded design export — strips everything), **Krea hosting FLUX 2** (no imwatermark despite FLUX — the host omits the encoder, same as Stability's hosted SDXL), and Midjourney (embeds nothing). Lesson: the imwatermark detector only fires on *pristine* output from a pipeline that runs the encoder (diffusers default, official BFL), not from re-hosts (Krea/Stability) or re-encoded exports (Recraft/Canva).
+- **Invisible but NOT locally detectable (proprietary, API/oracle only — same wall as SynthID):** Amazon Titan Image Generator + Nova Canvas (Bedrock `DetectGeneratedContent` API), Kakao (new SynthID image adopter, May 2026), NVIDIA Cosmos (SynthID video). No local detector possible; treat like SynthID.
+
 ## Known limitations

 - `invisible` pipeline downscales to model-native resolution (1024 px for SDXL) before diffusion. Degrades fine text in infographics. Tracked; fix is tile-based diffusion.
- Pyright first run is slow (2-3 min) due to ML deps (torch/diffusers/transformers stubs)
+- Pyright first run is slow (2-3 min) due to ML deps (torch/diffusers/transformers stubs); full-project `uv run pyright` can stall for many minutes — scope it to changed files.
 - `ultralytics` monkey-patches `PIL.Image.open` and tries to autoload `pi_heif`. When `pi_heif` is missing, opening files raises `ModuleNotFoundError`, not `UnidentifiedImageError`. Code that opens user-supplied or unknown-format files should `except Exception`, not just `OSError`/`UnidentifiedImageError`.
- Metadata detection for AVIF/HEIF/JPEG-XL relies on a binary scan for `C2PA_UUID` + `IPTC_AI_MARKERS`. C2PA removal in those containers is implemented via `noai/isobmff.py` (top-level ``uuid`` / ``jumb`` box stripper, no re-encoding). EXIF/XMP boxes inside those containers are not yet scrubbed.
- **SynthID v2 vs default pipeline:** the SDXL-based default profile (since May 2026) defeats SynthID v2 per the deployment in raiw-app (`fal-ai/fast-sdxl` at native ~1024 px, strength 0.05, steps 50). SD-1.5 dreamshaper at 768 px was previously the default and does NOT defeat v2 — verified empirically against Gemini app's "Verify with SynthID" feature (strength 0.04, 0.10, and elastic warp α∈{5,8} all flagged positive). That SD-1.5 path was removed; only `default` (SDXL) and `ctrlregen` profiles remain.
+- Metadata detection for AVIF/HEIF/JPEG-XL relies on a binary scan for `C2PA_UUID` + `IPTC_AI_MARKERS`, plus EXIF `Software` / XMP `CreatorTool` generator tags via `metadata.exif_generator` (validated with synthesized AVIF/JPEG fixtures + an XMP raw-scan fixture). C2PA removal in those containers is implemented via `noai/isobmff.py` (top-level ``uuid`` / ``jumb`` box stripper, no re-encoding). EXIF/XMP boxes inside those containers are read for detection but not yet **scrubbed** on removal.
+- **SynthID detection is metadata-only.** There is no reliable *local* detector of the SynthID *pixel* watermark — Google's decoder is proprietary, no public spec or API (only a waitlisted portal). We detect SynthID by its C2PA companion (`synthid_source` / `SYNTHID_C2PA_ISSUERS`), which is reliable while the manifest is intact but says nothing once C2PA is stripped. **Surface-dependent blind spot (verified 2026-05-24):** the same Google model emits different metadata per surface -- the Gemini *app* wraps outputs in Google C2PA, but the *API/playground* (AI Studio, Nano Banana / gemini-2.5-flash-image) emits the SynthID *pixel* watermark (confirmed via the Gemini-app oracle) + the visible sparkle but **no C2PA/IPTC at all**, so `synthid_source` returns None despite SynthID being present. Only the pixel oracle or the visible-sparkle detector catches those. (Meta AI is another surface mismatch: it writes the IPTC `digitalSourceType=trainedAlgorithmicMedia` marker, not C2PA and not SynthID.) Google→SynthID is long-standing; OpenAI→SynthID is confirmed by OpenAI's Help Center (ChatGPT/Codex/API "include both C2PA metadata and SynthID watermarks", updated 2026-05-21) but time-gated (pre-rollout OpenAI images carry C2PA without SynthID), so the OpenAI verdict is hedged "likely". Oracles: Gemini app "Verify with SynthID" (Google), openai.com/verify (OpenAI). The spectral phase-coherence approach from `github.com/aloshdenny/reverse-SynthID` was evaluated (May 2026) and **does not work for real-content detection**: on its own shipped codebook + validation set, watermarked and cleaned images were indistinguishable (conf within noise, cleaned often higher); it only fires on pure-black 1024x1024 reference images at exact resolution (the controlled case it was calibrated on). The README's "90% / conf=0.91" reproduces only in that lab condition. Do not build a production detector on it; if revisited, it is experimental/diagnostic only and needs a per-resolution, per-model reference corpus. A from-scratch gpt-image pilot (2026-05-24) confirmed this independently: 5 independent solid-black gpt-image outputs share a near-identical fixed signature (pairwise residual correlation **0.92**, avg-template retains 97% energy), so the watermark/carrier IS strongly present and consistent on flat content — but the carrier frequencies extracted from it do NOT discriminate real content (carrier-to-random ratio: cleaned 1.86 > watermarked 1.53; a non-gpt-image image scored highest at 3.67). The signature drowns in content texture. Net: a perfectly consistent solid-color signature still yields no real-content pixel detector with magnitude/carrier methods. A corpus discrimination test (2026-05-24, `scripts/synthid_pixel_probe.py`, raw zero-mean residual NCC) independently re-confirms this: at matched resolution, SynthID positives do NOT cluster apart from negatives (within-Gemini 0.07; at 1024 px pos-vs-neg >= pos-vs-pos). The only high correlations were near-duplicate *content* (5 ChatGPT renders of one prompt at ~0.92, while a distinct ChatGPT image scored ~0 against them) — content, not a carrier. The probe is solid-fills-only and EXPERIMENTAL/DIAGNOSTIC; do not use it on real content. **Correction (deeper re-examination 2026-05-25):** the carrier IS real on solid fills — the earlier "no carrier" was a *method* artifact of using spatial / FFT-magnitude NCC, which can't see it. The carrier is a fixed *phase* at specific low frequencies, so the right metric is **per-bin phase coherence**. On 8 white `gemini-2.5-flash-image` fills (generated via the reverse-SynthID trick: identity-edit prompt "Recreate this image exactly as it is" on a synthetic pure-white PNG — this bypasses the recitation block that rejects text prompts for pure colors), phase coherence at the white carriers `(0,±7..±12,±20..±23)` = **0.86** vs **0.31** random; single-image leave-one-out phase-match **+0.83** vs real photos **-0.24**. (Black `2.5-flash` fills clip to std≈0 — SynthID can't push values below 0, so no carrier in black; the repo's dark carriers come from nano-banana-pro.) **But it does not generalize:** (a) carriers are model-version + resolution + color specific — the repo's v4 codebook (built for `gemini-3.1-flash-image-preview` + `nano-banana-pro-preview`) scores ~0.527 on my 2.5-flash white fills, indistinguishable from negatives (~0.50), i.e. carriers shift across model versions and need a per-model codebook; (b) on real content (30 `2.5-flash` images) the carrier collapses — set phase coherence at carriers 0.37 ≈ random 0.42, and the repo's v4 detector gives content 0.518 ≈ negatives 0.504 (no separation; a faint +0.24 single-image lean is likely a brightness confound). Net: the spectral/phase approach is a real *controlled-fill* characterizer, NOT an arbitrary-real-content detector, and is brittle to model version. Metadata proxy + visible sparkle + online oracles remain the ceiling for real content.
+- **External AI-vs-real classifier models are out of scope (decided 2026-05-24).** Generic HuggingFace detectors (`Organika/sdxl-detector` Swin Transformer, `umm-maybe/AI-image-detector`, and fine-tunes) exist and report ~0.98 on their *own* SDXL-vs-real validation sets, but they are per-generator and the model cards themselves note degraded accuracy off-distribution; they are untested on gpt-image / Gemini Nano Banana (the metadata-stripped surfaces we care about), and our own light SDXL pass would likely defeat them the same way it defeats SynthID. Detection here stays local + signal-based (metadata + visible sparkle); do not add a bundled classifier dependency.
+- **SynthID v2 vs default pipeline:** the SDXL-based default profile (since May 2026) defeats SynthID v2. **Verified end-to-end (May 2026):** local SDXL run on a Gemini 3 Pro output, checked via the Gemini app's "Verify with SynthID" feature, returned "no SynthID watermark detected". Also confirmed against **OpenAI's** SynthID (2026-05-23): a fresh ChatGPT/gpt-image output read "SynthID detected" on openai.com/verify before the local SDXL run and "SynthID not detected" after (corpus regression chain: pos `4ef377bd` -> cleaned `47188e88`). The same configuration is used in raiw-app production (`fal-ai/fast-sdxl` at native ~1024 px, strength 0.05, steps 50). SD-1.5 dreamshaper at 768 px was previously the default and does NOT defeat v2 — verified empirically against the same feature (strength 0.04, 0.10, and elastic warp α∈{5,8} all flagged positive). That SD-1.5 path was removed; only `default` (SDXL) and `ctrlregen` profiles remain.
@@ -135,6 +135,13 @@ After installation the `remove-ai-watermarks` command is available system-wide.
 > ```bash
 > pip install -e ".[gpu]"   # or: uv pip install -e ".[gpu]"
 > ```
+>
+> To let `identify` decode the open Stable Diffusion / SDXL / FLUX invisible
+> watermarks, install the `detect` extra (adds the `invisible-watermark` decoder):
+>
+> ```bash
+> pip install -e ".[detect]"   # or: uv pip install -e ".[detect]"
+> ```

 #### Invisible watermark removal

@@ -180,6 +187,15 @@ remove-ai-watermarks batch ./images/ --mode all
 #### Individual commands

 ```bash
+# Identify provenance: where an image was made + its watermark inventory.
+# Aggregates C2PA, IPTC "Made with AI", embedded SD/ComfyUI params, EXIF/XMP
+# generator tags (incl. inside AVIF/HEIF), the SynthID proxy, the visible Gemini
+# sparkle, and (with the [detect] extra) the open SD/SDXL/FLUX invisible
+# watermark into one verdict. Reports "unknown"
+# (never "clean") when no signal is found, since stripped metadata is not proof
+# of a clean origin. Add --json for machine-readable output.
+remove-ai-watermarks identify image.png
+
 # Visible watermark only (Gemini / Nano Banana sparkle) — fast, offline
 remove-ai-watermarks visible image.png -o clean.png

@@ -187,6 +203,10 @@ remove-ai-watermarks visible image.png -o clean.png
 remove-ai-watermarks invisible image.png -o clean.png --humanize 4.0

 # Check / strip AI metadata (C2PA, EXIF, "Made with AI" labels)
+# --check also flags SynthID-bearing sources: a C2PA manifest signed by
+# Google or OpenAI implies an invisible SynthID watermark in the pixels
+# (both vendors pair the two). Adobe Firefly / Microsoft sign C2PA without
+# SynthID, so they are reported as C2PA only.
 remove-ai-watermarks metadata image.png --check
 remove-ai-watermarks metadata image.png --remove

@@ -253,7 +273,11 @@ pip install certifi

 Tracked but not yet implemented:

- **SynthID-Image v2 automated regression test**. The default SDXL profile defeats v2 per manual checks against the [Gemini app](https://support.google.com/gemini/answer/16722517)'s "Verify with SynthID" feature on a Gemini 3 Pro output (May 2026). An automated end-to-end test would need either programmatic access to the [SynthID Detector portal](https://blog.google/innovation-and-ai/products/google-synthid-ai-content-detector/) (waitlist for media professionals and researchers) or an offline surrogate detector. Open.
+- **SynthID-Image v2 automated regression test**. The default SDXL profile defeats v2 per manual checks against the [Gemini app](https://support.google.com/gemini/answer/16722517)'s "Verify with SynthID" feature on a Gemini 3 Pro output (May 2026). An automated end-to-end test would need either programmatic access to the [SynthID Detector portal](https://blog.google/innovation-and-ai/products/google-synthid-ai-content-detector/) (waitlist for media professionals and researchers) or an offline surrogate detector. The spectral phase-coherence surrogate from [reverse-SynthID](https://github.com/aloshdenny/reverse-SynthID) was evaluated and does not separate watermarked from cleaned real-content images (it only fires on controlled solid-color references at exact resolution), so it is not a usable oracle. Open.
+- **Local SynthID *pixel* detector**. Not feasible today: Google's decoder is proprietary, and magnitude/carrier spectral methods do not separate real content (confirmed by three independent evaluations, including a from-scratch gpt-image pilot; see CLAUDE.md). Blocked on either (a) a programmatic generation path (OpenAI / Gemini API) to build a per-(model, resolution) labeled corpus at scale, or (b) a raw watermarked-output dataset. If data arrives, the next approach to try is a learned classifier on diverse content rather than a fixed carrier codebook.
+- **Grow the SynthID reference corpus** (`data/synthid_corpus/`) with oracle-labeled samples per model and resolution (Gemini app for Google, openai.com/verify for OpenAI). Prerequisite for any pixel-detector attempt and for an automated removal-regression set.
+- **Real non-PNG C2PA fixtures**. SynthID-source detection for JPEG / WebP / AVIF is currently covered only by synthetic byte blobs; replace with real vendor-emitted files to ground the binary-scan path.
+- **Maintenance debt**. Bump transitive `idna` to 3.15 (blocks `uv-secure`); clear strict-pyright debt in `remove_ai_metadata` / `cli.py` (untyped piexif / PIL / click / rich) so `maintain.sh` can finish green.
 - **AVIF / HEIF / JPEG-XL detection limits**. Removal strips top-level C2PA `uuid` and JUMBF `jumb` boxes. EXIF/XMP boxes inside these containers are not yet scrubbed (PNG and JPEG are fully covered).
 - **Video pipeline (`noai-video`)**: per-frame inpainting and tracking for Sora 2 dynamic logo, Veo 3.1 badge, Kling, Runway. Separate package, not folded into this repo.

@@ -0,0 +1,139 @@
+# SynthID reference corpus
+
+A locally-collected, labeled image corpus for SynthID work. Two downstream uses:
+
+1. **Per-resolution spectral codebook** for an experimental SynthID detector
+   (carrier frequencies are resolution-dependent, so labels must record the
+   exact native resolution).
+2. **Removal regression set** — verify that our pipeline turns a SynthID-positive
+   image into a negative one.
+
+There is no reliable local detector of the SynthID pixel watermark (Google's
+decoder is proprietary). The ground-truth label therefore comes from an
+external oracle, recorded per image in `verified_via` (see below).
+
+## Layout
+
+```
+data/synthid_corpus/
+  README.md        # this protocol (committed)
+  manifest.csv     # labels + provenance (committed, reviewable)
+  images/          # the actual files (gitignored, local-only)
+    pos/           # SynthID present
+    neg/           # SynthID absent
+    cleaned/       # our pipeline output from a pos image
+```
+
+Images are gitignored on purpose: the corpus is large, may contain personal or
+licensed content, and SynthID-positive outputs are best kept local. The
+`manifest.csv` (sha256 + labels + extracted metadata) is the durable artifact.
+
+## Verification levels (`verified_via`)
+
+Ground-truth quality, strongest first:
+
+- `gemini-app` — checked via the Gemini app "Verify with SynthID" feature. Gold standard for the pixel watermark (Google models).
+- `openai-verify` — checked via openai.com/verify (gold standard for OpenAI ChatGPT/Codex/API images).
+- `synthid-portal` — checked via Google's SynthID Detector portal.
+- `c2pa-metadata` — issuer-only proxy (Google/OpenAI C2PA manifest present). Weaker: the C2PA can be stripped while the pixel watermark remains.
+- `third-party` — label asserted by an external dataset, not independently verified.
+- `none` — unverified.
+
+Prefer `gemini-app` for any image that will train the codebook or gate a test.
+
+## What to collect
+
+For the **codebook** (per target resolution, e.g. 1024x1024, 1024x1536, 1536x2816):
+
+- 30-50+ SynthID-positive outputs per resolution (more is better; ~150-200 per
+  resolution materially improves carrier discovery).
+- At each target resolution, also a batch of **pure-black (#000000)** and
+  **pure-white (#FFFFFF)** fills generated by the SynthID model — these isolate
+  the content-independent carrier (the watermark is most of the signal there).
+
+For the **regression set**:
+
+- A handful of `pos` images, their `cleaned` counterparts (run through our
+  pipeline), and the cleaned re-verified via `gemini-app` (should read negative).
+- `neg` controls: non-AI photos and outputs from non-SynthID models (SD,
+  Midjourney, Firefly) verified negative.
+
+Avoid personal or identifiable content; the corpus stays local.
+
+## Ingesting
+
+Use `scripts/synthid_corpus.py` — it copies a file in, records its sha256,
+resolution, format, and C2PA issuer (via our own detector), and appends a row
+to `manifest.csv`:
+
+```bash
+uv run python scripts/synthid_corpus.py ingest path/to/*.png \
+    --label pos --source "Gemini app" --model gemini-3-pro \
+    --verified-via gemini-app --notes "1024x1024 batch"
+
+uv run python scripts/synthid_corpus.py status   # counts by label / resolution / verification
+```
+
+## Autonomous collection via Chrome MCP
+
+Generation can be driven through the browser (the account must be logged in):
+
+- **Gemini** (`gemini.google.com`): type `Create an image: <prompt>`, wait, hover the
+  result, click the download icon (top-right). Single, reliable click. Outputs
+  carry Google C2PA + SynthID. Occasionally the composer stalls in a
+  "generating" state -> start a New chat to reset.
+- **ChatGPT** (`chatgpt.com`): the UI download is flaky (the fullscreen viewer
+  races and can grab the previous image; the share-modal path works but is
+  multi-step). Reliable path is an in-page fetch of the rendered image, which
+  preserves the original bytes (C2PA intact, unlike a canvas re-encode):
+
+  ```js
+  // run in the ChatGPT tab via the browser MCP javascript tool
+  (async () => {
+    const imgs = [...document.querySelectorAll('img')].filter(i => i.naturalWidth >= 400);
+    const img = imgs[imgs.length - 1];                 // newest large image
+    const b = await (await fetch(img.currentSrc || img.src)).blob();
+    const a = document.createElement('a');
+    a.href = URL.createObjectURL(b); a.download = 'dl.png';
+    document.body.appendChild(a); a.click(); a.remove();
+    return 'size=' + b.size;                            // do NOT return the src (privacy guard blocks query strings)
+  })()
+  ```
+
+  Gotcha: confirm the returned `size` differs from the previous image before
+  ingesting -- if the new image has not finished rendering, the script grabs the
+  prior one (the corpus dedups by sha256, but the notes would mislabel it).
+  ChatGPT also shows an A/B "which is better?" picker; click Skip first.
+
+**Originals, not previews.** Some platforms render a low-res preview in the chat
+(Grok serves a ~20KB 1024px JPEG/PNG; the in-page `<img>` fetch grabs *that*, not
+the original). Previews are re-encoded and **strip metadata**, so a "clean"
+preview is not proof the original is clean. Always pull the original via the
+platform's native Download / lightbox button and sanity-check the file size (a
+20KB "1024x1024" is a preview). ChatGPT's in-chat `<img>` *is* the full-res
+oaiusercontent original, so fetch+blob is fine there; Grok needs the lightbox
+Download; Leonardo serves the original JPEG in-chat (download button matches).
+
+## Per-platform watermark map (observed, May 2026)
+
+What each platform actually embeds, verified by byte-scan (and Gemini-app oracle
+where noted). The detector's coverage is complementary: metadata catches C2PA /
+IPTC; the visible detector catches the Gemini-family sparkle; the SynthID pixel
+itself has no local detector (oracle only).
+
+| Platform | C2PA issuer | SynthID pixel | IPTC "Made with AI" | Visible sparkle | Corpus label |
+|---|---|---|---|---|---|
+| Gemini app | Google | yes | - | yes | pos |
+| ChatGPT / gpt-image | OpenAI | yes | - | - | pos |
+| Microsoft Designer | OpenAI + Microsoft | yes (via OpenAI) | - | - | pos |
+| Google AI Studio (Nano Banana) | **none** | yes (oracle-confirmed) | - | yes | pos (metadata blind spot) |
+| Meta AI | none | no | **yes** | - | neg (for SynthID) |
+| Leonardo.ai | none | no | no | - | neg |
+| Grok (xAI) | none (non-adopter) | no | no | - | neg (captured: clean low-res preview) |
+| Bing Image Creator (DALL-E) | (expect OpenAI, like Designer) | - | - | - | not captured (UI uncooperative) |
+
+Key takeaways:
+- The same model differs by *surface*: Gemini app wraps C2PA, AI Studio (API/playground) emits none -- only the pixel + sparkle survive.
+- Microsoft Designer's DALL-E backend inherits OpenAI's C2PA+SynthID (issuer "OpenAI, Microsoft").
+- Meta uses the IPTC `digitalSourceType` marker, not C2PA or SynthID.
+- Bing and Grok web UIs are uncooperative for autonomous capture (no document_idle for screenshots; blob downloads intermittently no-op; low-res in-chat previews). Use their native download button manually if a full-res sample is needed.
@@ -0,0 +1,118 @@
+sha256,filename,label,source,model,width,height,format,c2pa_issuer,synthid_metadata,verified_via,added,notes
+4ef377bde1a1d4eff141972841938643b173f5052992a018b9a21b31ac31731e,"4ef377bd-ChatGPT Image May 23, 2026, 02_43_02 PM.png",pos,ChatGPT,gpt-image,1254,1254,png,OpenAI,yes,openai-verify,2026-05-23T21:48:12Z,fresh post-rollout 2026-05-23; openai.com/verify: SynthID+C2PA detected
+d09f84c0e4c6d8b336bf4a9a7277314e940dcb5052ae7051e785cbb3bb42d656,d09f84c0-Gemini_Generated_Image_vq7wkwvq7wkwvq7w.png,pos,Gemini app,gemini,2816,1536,png,Google LLC,yes,c2pa-metadata,2026-05-23T21:52:40Z,"user: latest Gemini, SynthID v2"
+47188e88f956291bd38ab6906e5f21eb273d4a697ddc8b4479deac9f48915e1a,47188e88-disco_synthid_removed.png,cleaned,our pipeline (invisible/SDXL),stabilityai/stable-diffusion-xl-base-1.0,1254,1254,png,,,openai-verify,2026-05-23T22:06:54Z,cleaned from 4ef377bd disco; openai.com/verify: SynthID NOT detected (defeated)
+52bb6bd524a74bff1b74ed893784c1c0ee76f48a9b45a009bae95ad4a57b7759,"52bb6bd5-ChatGPT Image May 24, 2026, 11_01_02 AM.png",pos,ChatGPT,gpt-image,1254,1254,png,OpenAI,yes,c2pa-metadata,2026-05-24T18:49:35Z,solid black ref (spectral pilot 2026-05-24)
+ee75ade07914d6306eae3443b3028782e0ac8c125a31b2d5a141f75aebdafb18,"ee75ade0-ChatGPT Image May 24, 2026, 11_10_04 AM.png",pos,ChatGPT,gpt-image,1254,1254,png,OpenAI,yes,c2pa-metadata,2026-05-24T18:49:35Z,solid black ref (spectral pilot 2026-05-24)
+9398c74dfad0f030633bd3ac224ba53e56a7ff9711d7b3a4c464e0073ece51b5,"9398c74d-ChatGPT Image May 24, 2026, 11_10_09 AM.png",pos,ChatGPT,gpt-image,1254,1254,png,OpenAI,yes,c2pa-metadata,2026-05-24T18:49:35Z,solid black ref (spectral pilot 2026-05-24)
+79c3733895e82e3c9e506de0ddd6dfbf20ba09171263d88444520777151868c8,"79c37338-ChatGPT Image May 24, 2026, 11_10_13 AM.png",pos,ChatGPT,gpt-image,1254,1254,png,OpenAI,yes,c2pa-metadata,2026-05-24T18:49:35Z,solid black ref (spectral pilot 2026-05-24)
+38315a0a83aa0e094a50520ea44e01aa26115927f09a532ff2ab8636de743e0a,"38315a0a-ChatGPT Image May 24, 2026, 11_10_19 AM.png",pos,ChatGPT,gpt-image,1254,1254,png,OpenAI,yes,c2pa-metadata,2026-05-24T18:49:35Z,solid black ref (spectral pilot 2026-05-24)
+f3a1fbc3bc8f768265400724bb9800d322f8e0b1461b2c585540845ea8352c5d,f3a1fbc3-winter_scene_X.png,neg,local (SDXL/processed),sdxl-or-processed,2816,1536,png,,,none,2026-05-24T18:52:23Z,metadata-clean: no C2PA/SynthID source
+f07bc0bcad09a5a5687ae312a1298c9ddd110a5e414265efc52ef4d524b36f86,f07bc0bc-api_enterprise_gate_ax.png,neg,local (SDXL/processed),sdxl-or-processed,1080,1350,png,,,none,2026-05-24T18:52:23Z,metadata-clean: no C2PA/SynthID source
+89571987e368f1ce82f9dedfa9101584434def842e50d1f4c759de64db5c21d9,89571987-c87bd3c48a4443a68cb84a65604dacd3_clean.png,neg,local (SDXL/processed),sdxl-or-processed,2816,1536,png,,,none,2026-05-24T18:52:23Z,metadata-clean: no C2PA/SynthID source
+7b650522d42db09568e249c04d683c469fb3e280a2c53fcd1031cb9df27c619a,"7b650522-ChatGPT Image May 24, 2026, 12_19_54 PM.png",pos,ChatGPT,gpt-image,1602,982,png,OpenAI,yes,c2pa-metadata,2026-05-24T19:20:25Z,content: misty pine forest at dawn
+fb28dba2a82cc101a92fdee5714867b32610d0564f37737fe4bb70782b8ecf32,fb28dba2-Gemini_Generated_Image_dsjlnsdsjlnsdsjl.png,pos,Gemini app,gemini,2816,1536,png,Google LLC,yes,c2pa-metadata,2026-05-24T19:30:25Z,content: elderly fisherman portrait
+d20d4cc936dbdfe909c52502039a9e84ba93d97b42b24a0acee5b7d6c71930ae,d20d4cc9-Gemini_Generated_Image_ug6kdpug6kdpug6k.png,pos,Gemini app,gemini,2816,1536,png,Google LLC,yes,c2pa-metadata,2026-05-24T19:33:15Z,content: red coffee mug product shot
+28f323345f6496d936c3f1a72f671ddf59d0f81565c24a63bf3286860f633afe,28f32334-chatgpt_fisherman.png,pos,ChatGPT,gpt-image,1023,1537,png,OpenAI,yes,c2pa-metadata,2026-05-24T19:39:20Z,content: elderly fisherman portrait (fetch+blob dl)
+88e61a384c2e0b12d97bc66046e4a10542b2987448ba89c4b49e66311e969c84,88e61a38-chatgpt_tokyo.png,pos,ChatGPT,gpt-image,1023,1537,png,OpenAI,yes,c2pa-metadata,2026-05-24T19:42:02Z,content: tokyo street night (fetch+blob dl)
+607087e40da1ac2636f7233c06784295d374d503ab928bb0c0c9b09e0140b3bc,607087e4-IMG_0505.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+1fa5f77f710c11a4cc69fed60195450def734401ed57c2600a84ce191f985440,1fa5f77f-IMG_0786.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+ac32346771d3ebb6b9d7d395e1efad995972fb708b76082905d15299d61dcfaa,ac323467-IMG_0114.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+2ea228ed270cd169e9d38bc3f1a162de7edbf54b6e5ad3c701a3c90010ec7067,2ea228ed-IMG_1791.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+eb0e14f7c8dfe88d710804fe0d2e74eb033d4d72db5525a526bb3751f2d52f4f,eb0e14f7-ai_price_software_not_hardware.png,neg,iPhone (me/post_photos),,1200,675,png,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+cd2c0eba3625bf7a90178063b4a769667cf00904db968efff7502734f8456892,cd2c0eba-IMG_3201.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+208b21ccfa8c81d2e75903a6083f2806ac068d856ad12957e6f6ea7870af0096,208b21cc-IMG_0712.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+f9c64806fd2b1ea4cf39a49f67e651f2146e9c892f1a401d9981e9d7f0601062,f9c64806-grep_vs_vector_agentic_search.png,neg,iPhone (me/post_photos),,1600,900,png,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+beaae48cf40c8d471c9c1a49d205cd87cf6c079211847f66ced4d4310aa7cff4,beaae48c-IMG_1699.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+41307ceeb9c981fa766ec4d711fe72f647fce86d21f1026e63c45dcd95063a65,41307cee-IMG_2148.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+3801fbe36a77134336508f888f658deb03657c5b765c8b3205b2d928e1b3a5f8,3801fbe3-IMG_2364.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+3012ff6d1db0535423c4dfb22e8d4d2da77b94b776c38cf06fa06958b7a71860,3012ff6d-IMG_2489.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+b978efbdec5ec96b320e53fb23a1aec306f5a32a796a56a032ae66929c803e7b,b978efbd-IMG_0334.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+1136f4df74e842e4ba95c91731656e4bf125361b0aa83b1fc4e128161668a8fe,1136f4df-IMG_0363.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+6e768835f52e840ec0f6af4a9604bd31ed38b2ce3fa187cfe5cc7daa2acb56e4,6e768835-IMG_3492.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+5470ded7892073b1deb31980b3cc541ce73b4cf025d56a43eb6e80c2db810ff2,5470ded7-IMG_1417.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+0bb4c176d83fbcbf4e628de7587d6183b9134c4f3fa85f55b96e94185273c7f6,0bb4c176-IMG_1790.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+c1f3267973e33e91c82ab889d21b6be18a6128e2178a85403bf89b04a6ea8c36,c1f32679-IMG_2981.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+efcb985a4ecc055029bea0df82ba161442340b1df6aa54639c2c717f8ebe7c4e,efcb985a-IMG_0841.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+383c6d9f4d74d1d30eeab2bfbe9b83212dcedb52b36b8ac403efa45730563326,383c6d9f-IMG_1616.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+b284d9ad7f58a44dc2b0d2e69b12236cf004d5052fb5484cfa1c21548bfd3821,b284d9ad-IMG_1250.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+85802ce2d2d8a7a155e631ef36a82cb912691fbbab4b2c28bb2438efed0b9225,85802ce2-IMG_2973.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+15b1e4e26340630345fc18e439ec98995ddca2816f38e13d4b2ee9e99ed26bac,15b1e4e2-IMG_1563.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+d8a1eed3cc327b5db7f5fd635fd4c37b29ff24a3360fce57622ea1cf35af284a,d8a1eed3-IMG_1666.heic,neg,iPhone (me/post_photos),,3024,4032,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+505047f592db5fd2ce4fc0fc3580e0b28abc19b6720a98a73530f2cb7158f7c1,505047f5-IMG_2569.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:47Z,post photo; no C2PA/SynthID (verified)
+a499773829f0ed206f9bd8c46740fc01955c1d43f45e8e775582015d9db9956c,a4997738-IMG_1980.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+a2f98b6a8887c7b4909601c46d03fb1f6e5145e714ffe3bc8b7056bc636bfee0,a2f98b6a-IMG_0780.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+9172987b14366231857e74a3f7b554ec0fc8cc23ebb7f88cd8eca3ce95ef8498,9172987b-IMG_3293.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+2f21fcd7b1ace88aa368028a7a97ffadd6ebe7cffe6c221d47abbed2a205074b,2f21fcd7-IMG_3144.PNG,neg,iPhone (me/post_photos),,2250,3000,png,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+d95e9af4d343501c4597918b5dfc66a1afc338b126e29c9eba13e7b2a61174d5,d95e9af4-IMG_1390.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+25b72de3a2ba4110c5dea68e82af6d19592eecc7000b2f12f52f8512afd48e73,25b72de3-IMG_1369.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+d18874c51b17ced8f736e24cd12ba54bd6c4f76292cba702a9c8a2d97c72f1c4,d18874c5-IMG_1447.HEIC,neg,iPhone (me/post_photos),,3088,2316,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+bf5abcd01037f44b2418b586c3b864340188a00f95047433832972ab797ae2d5,bf5abcd0-IMG_0743.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+a2782e0e6bb3d611925cace6ca3da10259ae70229aad3f366bc25b9e94538ab4,a2782e0e-IMG_0345.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+3fc0c4253f86427777904f41355eacbdcc1a29f6897ec694c7fc68b6e7f70846,3fc0c425-IMG_1450.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+56023f628e6d8ad0a3fb204b5bd10c9e9182446bad9032b0ab8c8d068a311bc7,56023f62-IMG_1796.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+100a121953049a83a714cd8771c7ac3f5afba69d8f8abb073535e98358395014,100a1219-IMG_0037.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+3e287f9e838f4f1bef5d8417957ea504954e5851de61656bbe32915c90db3a1f,3e287f9e-IMG_2323.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+237ff92d14163216687649f1fe222d5427f8600d8a97cb667386e71466a5d908,237ff92d-IMG_2270.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+a3f4a1a85bbd1cc6236e20c5a7d58bc18ccf695bc471cc02f36c1ad885551df8,a3f4a1a8-IMG_1309.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+be2f8bcb71f4bc5a4e79821bb4691e0f3c817ce916a35b05ffacce003a380c1c,be2f8bcb-IMG_0999.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+d5a4f2f81e86696cd18d38796c6b92ce96dcfe073bfd3a9c5dedf9d574aa66d6,d5a4f2f8-span_vs_flattening.png,neg,iPhone (me/post_photos),,1600,1000,png,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+2b148db2b1a314a87647a03828b3d235e7c4c939252f448b572b7658c7bb9723,2b148db2-IMG_1832.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+2d02fed0b6d60ce1142eadac9837a83896569dddbdad7cc5cfdec018f0506d36,2d02fed0-IMG_1520.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+68bf936a5491478484a295b3d882fef042f77e9784cae1b64a9e6e1b72d0628c,68bf936a-IMG_1107.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+32a9b71a13ecec3c90dcd7b54b90206acbe0f646a2296c955e227baa83823eec,32a9b71a-IMG_1238.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+8f170c06f843c2bcf4cf6e249dd76365287fdb3a322637ba93c570f50cb19772,8f170c06-IMG_1078.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+f7634a3779812f52de0599736b4c35f55b61027bae191e4ca060b10341661731,f7634a37-IMG_1605.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+e2ae859a96a5076f37b6efc5a1dad4952e29550f93cee4bba4899ef6fda09305,e2ae859a-IMG_3394.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+b4ea1e6f9134dfc91b4334ed723a15c937eee4fe3fd65b210cc31959884e5c4e,b4ea1e6f-IMG_1595.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+bf9d4d9ebb17cbae46407446d440b273fc136ea3e157a648e4959b116fca376d,bf9d4d9e-IMG_0916.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+2546689c376f365188a5e5a85e72982ff099b5680608e861c8490bde340e8a29,2546689c-IMG_2481.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+88944f46e3d85ea5c4f6d4559f8da9360f68060f73bb8b8c3344c3fdd3b54476,88944f46-IMG_2157.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+3fc4c8316012abc462df5534d4d995e5af84d35076a2927331107ff994293d9e,3fc4c831-IMG_2566.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+af59d9a72a859a72b6ad8e6178992e320934f74eb2a88d74c9d4c64b5bdb89f5,af59d9a7-IMG_2198.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+4f9d6807ff42d0bcd9494dc5d4c9d15f7b96691e29bfe2d670b8e3bceba934de,4f9d6807-IMG_0371.HEIC,neg,iPhone (me/post_photos),,3088,2316,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+237492e65e73acacf8135a2a903a2a231ee9c68ddcc2c590e90ec6e54c6a4520,237492e6-IMG_0473.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+d8a339db6b19c71d4ccf1a467c8999c5d2947b5c8cfd07a8b17b343f579e3f26,d8a339db-IMG_1798.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+abe73f771fb57c959663070cb560cfd5365999ef6b9aa419b18ea1031fab4a9d,abe73f77-IMG_3434.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+15fbe3e101bce529ec9b7a49fb668ecdad62c3aceb6e649f4668b58af1784c32,15fbe3e1-IMG_0481.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+455d618e94ebf2e469ede016e06c48def5cc14ef68c090584b31570b9fc5e34e,455d618e-IMG_2957.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+dae46cfca7baa921675ffaf74701617b6fe10e629c5e7f099c67bf21a3f2a62c,dae46cfc-IMG_0356.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+cb4d61e66982736d552b043866d80bfb304fe1f44ec07f3acabeff71d08b3708,cb4d61e6-IMG_1681.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+c7776595d517bebab73370b6395deb1dc2bb507f99d2bd551bfe1a2416f5165b,c7776595-168E1031-ADC1-48A2-8B63-F5EB228DF5B4.JPG,neg,iPhone (me/post_photos),,3664,2062,jpeg,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+5fe59521d579b536340253d2bcaa7c28ddd2485fa964fc27a9b9e5118ad0cdd1,5fe59521-IMG_1496.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+81e4e532c124b182b1608d19feda52b829390adc9b77f47b72d89829035917ff,81e4e532-IMG_2193.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+4ead7918a5aeea7fba78b36af44358fea7ed1db7f46f4bc675152b9d04e68c38,4ead7918-IMG_1300.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:48Z,post photo; no C2PA/SynthID (verified)
+75f8a25c217374689d8224d58736f09245882334a9e4d39931025bd66a6df341,75f8a25c-IMG_3540.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+574acd7b8a49652d62df90539c2e18d1f3f510399a294f6c5209ab725c96a7df,574acd7b-IMG_3393.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+8115feaa7b758fd1b984c55355fc8f4c10dfe02192fe810bb6692556190d5176,8115feaa-IMG_1341.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+4c795238b89178cf52a3674e721ed7f4cd5068028385491d12171a3c62545c35,4c795238-IMG_3034.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+5abfaccb37c549de67c7f3a751a528a423e24a82c643fdb29094be8debbb206d,5abfaccb-IMG_3018.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+5fed1923d513c1e9ffcba2f240e617fa9344fb39d35144169570ece8b0bd0f33,5fed1923-IMG_0272.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+b3e1b034bc9ff1fd7e3db77ae0f0a695fcb4ed7bdcb473312427d78aa18700a1,b3e1b034-IMG_1199.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+79a12ccab0ccb954b1917d8edc75cb7c01d02c3a803b34f45be7afdb69cf80ea,79a12cca-IMG_0767.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+7252adc2103f7be6d04d56e0eed0d90122c8203af0ae08c28f922338b7c1bbb5,7252adc2-IMG_0623.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+06b04d8fe8e1cd6bee9a973f93bfda37586924cbfec7d372f59d52aa9196160b,06b04d8f-IMG_1474.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+591636c01c57864c44bdf7c1062bf6fca87efa8fae25bfc58c7cffb5a73424ab,591636c0-IMG_2809.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+ca2f226f45a5e37071605f074fa64de31cd09f923b22f57bb8255aa9046c6b92,ca2f226f-IMG_1419.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+ed751e2489ce4b26c268019942ed65a1f95a10f8056f26ac646d7d19e5db58ad,ed751e24-IMG_1225.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+3757e6100e567cbbf8eadf4cb8a064d79073cea1588251125a3b1ae13893e778,3757e610-IMG_2583.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+e973424c779cfa5dcd2403f103673423771192509395586f8efc459c92240d5d,e973424c-IMG_2215.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+46657fb69680b3e5db8fdb17aa306f2b6ee41a719408ac5812ff4cb399e15485,46657fb6-IMG_1439.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+dd112d116dda31847b50f40ab0990306d1c1f137780c67bb094ee6992f2d6f7e,dd112d11-IMG_1317.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+a0cdcc17d39eda2a7b2c34ff1f4ed97290fa996fef530c1d24da91722b1a6893,a0cdcc17-IMG_0005.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+70fabcbeedafb3c033d5805328ffd12bf644102e8666b878eca5968bd075b3fe,70fabcbe-IMG_3500.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+8fdb574a94e65e14ac29017cf2d5a2ede18a8c4e3f12e04c64292b0d38570062,8fdb574a-IMG_3557.HEIC,neg,iPhone (me/post_photos),,4032,3024,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+cb6777332e357aa38916400c2c1fe76801aad54946e789e9fa6a7d4659d02948,cb677733-IMG_3582.HEIC,neg,iPhone (me/post_photos),,5712,4284,heic,,,none,2026-05-24T20:58:49Z,post photo; no C2PA/SynthID (verified)
+72c71846fdb9eb3d960142c24627304349ee4fd8665d10d06e336430b0e20d80,72c71846-meta_lake.jpg,neg,Meta AI,meta,810,1440,jpeg,,,none,2026-05-24T21:33:47Z,"IPTC Made-with-AI (trainedAlgorithmicMedia); NO SynthID, NO C2PA"
+c86973424817f62510e2a312b85c52e05adf47ace87a8e717fd442607596f501,c8697342-aistudio_lake.png,pos,Google AI Studio (Nano Banana),gemini-2.5-flash-image,1024,1024,png,,,gemini-app,2026-05-24T21:39:09Z,"API/playground: SynthID pixel CONFIRMED (Gemini-app oracle) + visible sparkle, but NO C2PA/IPTC -> synthid_source blind spot"
+1f81827c06d67cf6f6c7f5d53ec8f9738183942a6d1d2717b161fea0fdcc540a,1f81827c-Designer.png,pos,Microsoft Designer,dall-e (Designer),1024,1024,png,"OpenAI, Microsoft",yes,c2pa-metadata,2026-05-24T22:18:40Z,C2PA issuer OpenAI+Microsoft; synthid_source=OpenAI (DALL-E surface inherits OpenAI SynthID+C2PA)
+f3ebe4683fc40aa2a0b80cc3ced3dc8062385ae32b488f4b33cb2330949e658d,f3ebe468-leonardo_lake.jpg,neg,Leonardo.ai,lucid-origin,1024,1024,jpeg,,,none,2026-05-24T22:22:15Z,non-SynthID AI; NO provenance metadata (no C2PA/IPTC/SynthID)
+12775f4c0aec1ecd615e88c4941a788a053d0258d08d31ad34d330d3e21969bd,12775f4c-grok_original.jpg,neg,Grok (xAI),aurora,784,1168,jpeg,,,none,2026-05-24T22:47:46Z,full-res original via Download button; xAI non-adopter; NO C2PA/IPTC/SynthID (has benign EXIF block)
+74b605deb102d74443a182f70fbea34a6a470bed381d62ae0bfa0083a6ecc5a3,74b605de-_e1f7984d-98e7-4cdc-91d2-f72314a4bc56.jpeg,neg,Bing Image Creator,MAI-Image-2e,1024,1024,jpeg,,,c2pa-metadata,2026-05-25T00:10:28Z,C2PA issuer Microsoft; non-SynthID; issuer-coverage fixture
+f9bb6b1039ee228221d2eead7daebd46422a36a4a8959a74fa0198e1aa19f9b8,f9bb6b10-a_red_ceramic_coffee_mug_on_a_wooden_table_soft_morning_light_photorealistic_20260525000725_04.png,neg,Stability Brand Studio,Stable Image,1024,1024,png,Stability AI,,c2pa-metadata,2026-05-25T00:10:31Z,C2PA issuer Stability AI; non-SynthID; no imwatermark
+23a06e0067930a9ad09a803f634414eb8ca0234fc0bbaca8d33f138745175e2e,23a06e00-a-red-ceramic-coffee-mug-on-a-wooden-table--soft-m.png,neg,Recraft V4.1,recraft-v4,896,1152,png,,,none,2026-05-25T01:21:35Z,"PNG export, no C2PA/EXIF/IPTC/imwatermark; clean download"
+4f3fea84c24fda8d72a6e7bfbae31827348d9edc755ec9a5e8c835860130de57,4f3fea84-a_red_ceramic_coffee_mug_on_a_wooden_table_soft_morning_light_photorealistic_67qgxtyin9kjk5yb5mph_0.png,neg,Krea (FLUX 2 host),flux-2-dev via Krea,832,1248,png,,,none,2026-05-25T01:21:38Z,no imwatermark despite FLUX; Krea host strips/omits; clean download
+7bdff2c3131cdc21783282a8a95c88f4b14ad38494a1041d0d68691af7fa5cd8,7bdff2c3-0001_1_a-red-ceramic-coffee-mug-sits-on-weather_D3BRZfOyUwuh5Emjtu8v4g_F7yI1GFGTayDA1NAUmkraw_cover.jpeg,neg,Ideogram,ideogram-3,1024,1024,jpeg,,,none,2026-05-25T01:36:38Z,EXIF Make='Ideogram AI'; no C2PA/SynthID; caught via exif_generator
+ca2162624607b8f030c63cb1371d6afdd3f0d7af3d9b17514f7ae64a2e8f378c,ca216262-0001_2_a-red-ceramic-coffee-mug-sits-on-weather_rs0Lt1beXV-Mwjs33vWKVw_F7yI1GFGTayDA1NAUmkraw.jpeg,neg,Ideogram,ideogram-3,1024,1024,jpeg,,,none,2026-05-25T01:36:38Z,EXIF Make='Ideogram AI'; no C2PA/SynthID; caught via exif_generator
+8a90f2ddf83ceedd7944531d27248191c79dc321ec76ea1f36df6cff9f6b8916,8a90f2dd-0001_3_a-red-ceramic-coffee-mug-sits-on-weather_Lm96lBwcXsmdeCfCpm52vw_F7yI1GFGTayDA1NAUmkraw.jpeg,neg,Ideogram,ideogram-3,1024,1024,jpeg,,,none,2026-05-25T01:36:38Z,EXIF Make='Ideogram AI'; no C2PA/SynthID; caught via exif_generator
+87fca198d5c436b27a16685972802a2ff0a310a941248c9ec8d81560d7f387a1,87fca198-0001_4_a-red-ceramic-coffee-mug-sits-on-weather_pGmlgcH4XqSaq41hc0a9AQ_F7yI1GFGTayDA1NAUmkraw.jpeg,neg,Ideogram,ideogram-3,1024,1024,jpeg,,,none,2026-05-25T01:36:38Z,EXIF Make='Ideogram AI'; no C2PA/SynthID; caught via exif_generator
@@ -1,6 +1,6 @@
 [project]
 name = "remove-ai-watermarks"
-version = "0.4.1"
+version = "0.5.0"
 description = "Remove visible and invisible AI watermarks from images (Gemini / Nano Banana, ChatGPT, Stable Diffusion)"
 readme = "README.md"
 requires-python = ">=3.10"
@@ -26,13 +26,21 @@ gpu = [
    "ultralytics>=8.0.0",
    "color-matcher>=0.5.0",
 ]
+# Open invisible-watermark (imwatermark) decoder for detecting the DWT-DCT
+# watermarks embedded by Stable Diffusion / SDXL / FLUX. Optional because it
+# pulls non-headless opencv; identify() guards the import and skips the signal
+# when absent.
+detect = [
+    "invisible-watermark>=0.2.0",
+]
 dev = [
    "pytest>=8.0.0",
    "pytest-cov>=4.1.0",
    "ruff>=0.4.0",
    "pyright>=1.1.0",
+    "invisible-watermark>=0.2.0",
 ]
-all = ["remove-ai-watermarks[gpu,dev]"]
+all = ["remove-ai-watermarks[gpu,detect,dev]"]

 # diffusers 0.38.0 (security fix for GHSA-98h9-4798-4q5v) declares a dependency
 # on safetensors>=0.8.0rc0 — a pre-release. Allow pre-releases globally so the
@@ -0,0 +1,240 @@
+"""Ingest and inspect the local SynthID reference corpus.
+
+Copies images into ``data/synthid_corpus/images/<label>/`` and records one row
+per image in ``manifest.csv`` (sha256, resolution, format, C2PA issuer, and the
+external verification level). Dogfoods the project's own C2PA detector so the
+recorded metadata matches what the library sees.
+
+See ``data/synthid_corpus/README.md`` for the collection protocol.
+
+Usage:
+    uv run python scripts/synthid_corpus.py ingest IMAGES... --label pos \\
+        --source "Gemini app" --model gemini-3-pro --verified-via gemini-app
+    uv run python scripts/synthid_corpus.py status
+"""
+
+from __future__ import annotations
+
+import csv
+import hashlib
+import logging
+import shutil
+import subprocess
+import sys
+from collections import Counter
+from datetime import datetime, timezone
+from pathlib import Path
+
+import click
+from PIL import Image
+from rich.console import Console
+from rich.table import Table
+
+from remove_ai_watermarks.noai.c2pa import extract_c2pa_info
+
+log = logging.getLogger(__name__)
+console = Console()
+
+DEFAULT_ROOT = Path(__file__).resolve().parent.parent / "data" / "synthid_corpus"
+LABELS = ("pos", "neg", "cleaned")
+VERIFIED_VIA = ("gemini-app", "openai-verify", "synthid-portal", "c2pa-metadata", "third-party", "none")
+FIELDNAMES = [
+    "sha256",
+    "filename",
+    "label",
+    "source",
+    "model",
+    "width",
+    "height",
+    "format",
+    "c2pa_issuer",
+    "synthid_metadata",
+    "verified_via",
+    "added",
+    "notes",
+]
+
+
+def _manifest_path(root: Path) -> Path:
+    return root / "manifest.csv"
+
+
+def _read_manifest(root: Path) -> list[dict[str, str]]:
+    path = _manifest_path(root)
+    if not path.exists():
+        return []
+    with open(path, newline="") as f:
+        return list(csv.DictReader(f))
+
+
+def _sha256(path: Path) -> str:
+    h = hashlib.sha256()
+    with open(path, "rb") as f:
+        for chunk in iter(lambda: f.read(1 << 20), b""):
+            h.update(chunk)
+    return h.hexdigest()
+
+
+def _dims_via_sips(path: Path) -> tuple[int, int]:
+    """Read pixel dimensions via macOS ``sips`` (handles HEIC, which PIL can't
+    open without pillow-heif). Returns (0, 0) off macOS or on failure."""
+    sips = shutil.which("sips")
+    if sys.platform != "darwin" or not sips:
+        return 0, 0
+    try:
+        # Fixed command; the only variable is a local file path passed to ingest.
+        out = subprocess.run(  # noqa: S603
+            [sips, "-g", "pixelWidth", "-g", "pixelHeight", str(path)],
+            capture_output=True,
+            text=True,
+            timeout=15,
+        ).stdout
+        dims = {"pixelWidth": 0, "pixelHeight": 0}
+        for line in out.splitlines():
+            for key in dims:
+                if f"{key}:" in line:
+                    dims[key] = int(line.split(":")[1].strip())
+        return dims["pixelWidth"], dims["pixelHeight"]
+    except Exception:  # sips missing/odd output; dims stay unknown
+        return 0, 0
+
+
+def _probe(path: Path) -> tuple[int, int, str, str, bool]:
+    """Return (width, height, format, c2pa_issuer, synthid_metadata)."""
+    width = height = 0
+    fmt = path.suffix.lstrip(".").lower()
+    try:
+        with Image.open(path) as img:
+            width, height = img.size
+            fmt = (img.format or fmt).lower()
+    except Exception as exc:  # unknown/user formats can raise non-OSError; see CLAUDE.md
+        log.debug("PIL could not open %s: %s", path, exc)
+
+    if width == 0:  # e.g. HEIC from iPhone — fall back to macOS sips
+        width, height = _dims_via_sips(path)
+
+    info = extract_c2pa_info(path)
+    issuer = str(info.get("issuer", ""))
+    synthid = "synthid_watermark" in info
+    return width, height, fmt, issuer, synthid
+
+
+@click.group()
+def cli() -> None:
+    """Manage the local SynthID reference corpus."""
+    logging.basicConfig(level=logging.INFO, format="%(message)s")
+
+
+@cli.command()
+@click.argument("images", nargs=-1, required=True, type=click.Path(exists=True, path_type=Path))
+@click.option("--label", required=True, type=click.Choice(LABELS), help="SynthID label.")
+@click.option("--source", default="", help="Where the image came from (free text).")
+@click.option("--model", default="", help="Generating model, e.g. gemini-3-pro.")
+@click.option(
+    "--verified-via", "verified_via", default="none", type=click.Choice(VERIFIED_VIA), help="Ground-truth oracle."
+)
+@click.option("--notes", default="", help="Free-text notes (e.g. resolution batch).")
+@click.option("--root", type=click.Path(path_type=Path), default=DEFAULT_ROOT, help="Corpus root.")
+def ingest(
+    images: tuple[Path, ...],
+    label: str,
+    source: str,
+    model: str,
+    verified_via: str,
+    notes: str,
+    root: Path,
+) -> None:
+    """Copy IMAGES into the corpus and append rows to the manifest."""
+    dest_dir = root / "images" / label
+    dest_dir.mkdir(parents=True, exist_ok=True)
+
+    rows = _read_manifest(root)
+    seen = {r["sha256"] for r in rows}
+    added = 0
+    new_rows: list[dict[str, str]] = []
+
+    for src in images:
+        digest = _sha256(src)
+        if digest in seen:
+            console.print(f"  [dim]skip (duplicate sha256): {src.name}[/]")
+            continue
+        seen.add(digest)
+
+        width, height, fmt, issuer, synthid = _probe(src)
+        stored_name = f"{digest[:8]}-{src.name}"
+        shutil.copy2(src, dest_dir / stored_name)
+
+        new_rows.append(
+            {
+                "sha256": digest,
+                "filename": stored_name,
+                "label": label,
+                "source": source,
+                "model": model,
+                "width": str(width),
+                "height": str(height),
+                "format": fmt,
+                "c2pa_issuer": issuer,
+                "synthid_metadata": "yes" if synthid else "",
+                "verified_via": verified_via,
+                "added": datetime.now(timezone.utc).strftime("%Y-%m-%dT%H:%M:%SZ"),
+                "notes": notes,
+            }
+        )
+        added += 1
+        flag = " [yellow](C2PA-SynthID)[/]" if synthid else ""
+        console.print(f"  [green]+[/] {label}/{stored_name}  {width}x{height} {fmt}{flag}")
+
+    if new_rows:
+        path = _manifest_path(root)
+        write_header = not path.exists()
+        path.parent.mkdir(parents=True, exist_ok=True)
+        with open(path, "a", newline="") as f:
+            writer = csv.DictWriter(f, fieldnames=FIELDNAMES)
+            if write_header:
+                writer.writeheader()
+            writer.writerows(new_rows)
+
+    console.print(f"\n  Ingested [bold]{added}[/] image(s); skipped {len(images) - added} duplicate(s).")
+
+
+@cli.command()
+@click.option("--root", type=click.Path(path_type=Path), default=DEFAULT_ROOT, help="Corpus root.")
+def status(root: Path) -> None:
+    """Print corpus counts by label, resolution, and verification level."""
+    rows = _read_manifest(root)
+    if not rows:
+        console.print("  [dim]Corpus is empty.[/]")
+        return
+
+    by_label = Counter(r["label"] for r in rows)
+    by_verified = Counter(r["verified_via"] for r in rows)
+    by_res = Counter(f"{r['label']}  {r['width']}x{r['height']}" for r in rows)
+
+    console.print(f"\n  [bold]{len(rows)}[/] image(s) in {root}")
+
+    t = Table(title="By label", show_header=True, header_style="bold")
+    t.add_column("Label")
+    t.add_column("Count", justify="right")
+    for k in LABELS:
+        if by_label.get(k):
+            t.add_row(k, str(by_label[k]))
+    console.print(t)
+
+    t = Table(title="By label x resolution", show_header=True, header_style="bold")
+    t.add_column("Label / resolution")
+    t.add_column("Count", justify="right")
+    for k, v in sorted(by_res.items()):
+        t.add_row(k, str(v))
+    console.print(t)
+
+    t = Table(title="By verification", show_header=True, header_style="bold")
+    t.add_column("verified_via")
+    t.add_column("Count", justify="right")
+    for k, v in by_verified.most_common():
+        t.add_row(k, str(v))
+    console.print(t)
+
+
+if __name__ == "__main__":
+    cli()
@@ -0,0 +1,149 @@
+"""SynthID pixel-carrier probe -- EXPERIMENTAL / DIAGNOSTIC ONLY.
+
+There is no local detector of the SynthID pixel watermark on real content: the
+carrier drowns in scene texture (see CLAUDE.md, confirmed repeatedly). This
+probe is meaningful ONLY on **solid-color fills**, where the per-pixel deviation
+from the image mean is essentially the watermark carrier (almost all the
+variance). It answers two controlled questions, neither of which is a
+real-content detector:
+
+  consistency IMAGES...
+      Mean pairwise normalized cross-correlation (NCC) of the carriers across
+      independent solid fills from one model, vs a random baseline. Genuine
+      SynthID positives share a fixed carrier, so they correlate well above
+      random (the pilot saw ~0.92 on gpt-image black fills); clean fills don't.
+
+  removal --pos P... --cleaned C...
+      Build a carrier template from the positive fills, then compare how the
+      positives and the pipeline-cleaned fills correlate to it. If removal
+      worked, the cleaned correlation collapses toward the random baseline --
+      pixel-domain evidence that the pipeline destroys the carrier, not just the
+      C2PA metadata.
+
+Do NOT run this on real-content images; the numbers are uninformative there.
+"""
+
+from __future__ import annotations
+
+import logging
+from typing import TYPE_CHECKING
+
+import click
+import numpy as np
+from PIL import Image
+from rich.console import Console
+
+if TYPE_CHECKING:
+    from numpy.typing import NDArray
+
+log = logging.getLogger(__name__)
+console = Console()
+
+
+def load_gray(path: str) -> NDArray[np.float64]:
+    """Load an image as a float64 grayscale array (mean of RGB channels)."""
+    with Image.open(path) as img:
+        return np.asarray(img.convert("RGB"), dtype=np.float64).mean(axis=2)
+
+
+def carrier(gray: NDArray[np.float64]) -> NDArray[np.float64]:
+    """Zero-mean, unit-norm residual of a solid-fill image -- its carrier.
+
+    Returns a flattened unit-norm vector for NCC comparison. A perfectly flat
+    image (std 0, e.g. a synthetic #000000 reference) has no carrier and yields
+    an all-zero vector, which correlates to 0 with everything.
+    """
+    residual = gray - float(gray.mean())
+    norm = float(np.linalg.norm(residual))
+    if norm == 0.0:
+        return residual.ravel()
+    return (residual / norm).ravel()
+
+
+def ncc(a: NDArray[np.float64], b: NDArray[np.float64]) -> float:
+    """Normalized cross-correlation of two carriers (unit-norm zero-mean vectors)."""
+    if a.shape != b.shape or a.size == 0:
+        return 0.0
+    return float(np.dot(a, b))
+
+
+def mean_pairwise_ncc(carriers: list[NDArray[np.float64]]) -> float:
+    """Average NCC over all distinct carrier pairs; 0.0 if fewer than two."""
+    scores = [ncc(carriers[i], carriers[j]) for i in range(len(carriers)) for j in range(i + 1, len(carriers))]
+    return float(np.mean(scores)) if scores else 0.0
+
+
+def template(carriers: list[NDArray[np.float64]]) -> NDArray[np.float64]:
+    """Average carrier, renormalized to unit norm (the shared-pattern estimate)."""
+    avg = np.mean(carriers, axis=0)
+    norm = float(np.linalg.norm(avg))
+    return avg / norm if norm else avg
+
+
+def random_baseline(shape: tuple[int, ...], n: int, *, seed: int = 0) -> float:
+    """Mean pairwise NCC of ``n`` random-noise carriers of ``shape`` (~0)."""
+    rng = np.random.default_rng(seed)
+    noise = [carrier(rng.standard_normal(shape)) for _ in range(max(n, 2))]
+    return mean_pairwise_ncc(noise)
+
+
+def _load_carriers(paths: tuple[str, ...]) -> list[NDArray[np.float64]]:
+    """Load carriers for same-shaped images; warn and skip mismatched shapes."""
+    grays = [(p, load_gray(p)) for p in paths]
+    shape = grays[0][1].shape
+    carriers: list[NDArray[np.float64]] = []
+    for p, g in grays:
+        if g.shape != shape:
+            console.print(f"  [yellow]skip[/] {p}: shape {g.shape} != {shape}")
+            continue
+        carriers.append(carrier(g))
+    return carriers
+
+
+@click.group()
+def cli() -> None:
+    """SynthID pixel-carrier probe (solid-color fills only)."""
+
+
+@cli.command()
+@click.argument("images", nargs=-1, required=True, type=click.Path(exists=True))
+def consistency(images: tuple[str, ...]) -> None:
+    """Mean pairwise carrier NCC across solid fills, vs the random baseline."""
+    carriers = _load_carriers(images)
+    if len(carriers) < 2:
+        console.print("[red]Need at least two same-shaped images.[/]")
+        raise SystemExit(1)
+    observed = mean_pairwise_ncc(carriers)
+    baseline = random_baseline(carriers[0].shape, len(carriers))
+    console.print(f"  carriers:          {len(carriers)}")
+    console.print(f"  mean pairwise NCC: [bold]{observed:.3f}[/]")
+    console.print(f"  random baseline:   {baseline:.3f}")
+    verdict = "shared carrier present" if observed > 0.3 else "no shared carrier (within noise)"
+    console.print(f"  verdict: [bold]{verdict}[/]")
+
+
+@cli.command()
+@click.option("--pos", "pos", multiple=True, required=True, type=click.Path(exists=True), help="Positive solid fills.")
+@click.option(
+    "--cleaned", "cleaned", multiple=True, required=True, type=click.Path(exists=True), help="Pipeline-cleaned fills."
+)
+def removal(pos: tuple[str, ...], cleaned: tuple[str, ...]) -> None:
+    """Does the pipeline drop the carrier correlation toward the random baseline?"""
+    pos_carriers = _load_carriers(pos)
+    cleaned_carriers = _load_carriers(cleaned)
+    if not pos_carriers or not cleaned_carriers:
+        console.print("[red]Need at least one positive and one cleaned fill of matching shape.[/]")
+        raise SystemExit(1)
+    tmpl = template(pos_carriers)
+    pos_corr = float(np.mean([ncc(c, tmpl) for c in pos_carriers]))
+    cleaned_corr = float(np.mean([ncc(c, tmpl) for c in cleaned_carriers]))
+    baseline = random_baseline(tmpl.shape, max(len(cleaned_carriers), 2))
+    console.print(f"  positive->template NCC: [bold]{pos_corr:.3f}[/]")
+    console.print(f"  cleaned->template NCC:  [bold]{cleaned_corr:.3f}[/]")
+    console.print(f"  random baseline:        {baseline:.3f}")
+    effective = cleaned_corr < pos_corr / 2
+    console.print(f"  verdict: [bold]{'carrier attenuated' if effective else 'carrier survives'}[/]")
+
+
+if __name__ == "__main__":
+    cli()
@@ -1,3 +1,3 @@
 """Remove-AI-Watermarks: Unified tool for removing visible and invisible AI watermarks."""

-__version__ = "0.3.4"
+__version__ = "0.5.0"
@@ -8,6 +8,7 @@ Provides commands for:

 from __future__ import annotations

+import json
 import logging
 import time
 from pathlib import Path
@@ -310,6 +311,8 @@ def cmd_metadata(
        if has_ai:
            console.print(f"  [yellow]⚠[/] AI metadata detected in {source.name}:")
            meta = get_ai_metadata(source)
+            if synthid := meta.get("synthid_watermark"):
+                console.print(f"  [bold yellow]⚠ SynthID pixel watermark {synthid}[/]")
            table = Table(show_header=True, header_style="bold")
            table.add_column("Key", style="cyan")
            table.add_column("Value")
@@ -327,6 +330,59 @@ def cmd_metadata(
    console.print(f"  [green]✓[/] AI metadata stripped → {out}")


+# ── Provenance identification ───────────────────────────────────────
+
+
+@main.command("identify")
+@click.argument("source", type=click.Path(exists=True, path_type=Path))
+@click.option(
+    "--no-visible",
+    is_flag=True,
+    help="Skip pixel-domain detectors (visible sparkle + invisible watermark); metadata-only.",
+)
+@click.option("--json", "as_json", is_flag=True, help="Emit the report as JSON instead of a table.")
+@click.pass_context
+def cmd_identify(ctx: click.Context, source: Path, no_visible: bool, as_json: bool) -> None:
+    """Identify where an image was made and what watermarks it carries.
+
+    Aggregates C2PA Content Credentials, IPTC "Made with AI" tags, embedded
+    generation parameters, the SynthID metadata proxy, and the visible Gemini
+    sparkle into a single provenance verdict. Absence of signals is reported as
+    "unknown", never as "clean" (stripped metadata leaves no local proof).
+    """
+    from dataclasses import asdict
+
+    from remove_ai_watermarks.identify import identify
+
+    source = _validate_image(source)
+    report = identify(source, check_visible=not no_visible, check_invisible=not no_visible)
+
+    if as_json:
+        click.echo(json.dumps(asdict(report), default=str, indent=2))
+        return
+
+    _banner()
+    verdict = {True: "[yellow]AI-generated[/]", False: "[green]not AI[/]", None: "[dim]unknown[/]"}[
+        report.is_ai_generated
+    ]
+    console.print(f"\n  Verdict: {verdict}  [dim](confidence: {report.confidence})[/]")
+    console.print(f"  Platform: {report.platform or '[dim]undetermined[/]'}")
+
+    if report.watermarks:
+        table = Table(show_header=True, header_style="bold", title="Watermarks / provenance markers")
+        table.add_column("Marker", style="cyan")
+        for wm in report.watermarks:
+            table.add_row(wm)
+        console.print(table)
+    else:
+        console.print("  [dim]No watermarks or provenance markers found.[/]")
+
+    if report.caveats:
+        console.print("\n  [dim]Caveats:[/]")
+        for c in report.caveats:
+            console.print(f"  [dim]- {c}[/]")
+
+
 # ── Combined "all" mode ──────────────────────────────────────────────


@@ -677,7 +733,7 @@ def cmd_batch(
                    seed=seed,
                    hf_token=hf_token,
                    humanize=humanize,
-                    )
+                )
                processed += 1

            except Exception as e:
@@ -546,3 +546,17 @@ class GeminiEngine:
            result[py1:py2, px1:px2] = blended.astype(np.uint8)

        return result
+
+
+def detect_sparkle_confidence(image_path: Path) -> float | None:
+    """Visible-sparkle detection confidence for a file, for provenance use.
+
+    Loads the image with cv2 and runs :meth:`GeminiEngine.detect_watermark`.
+    Returns the NCC confidence in [0, 1], or None if the image cannot be read
+    (cv2 returns None for unsupported containers such as HEIC). Kept here so the
+    cv2 dependency stays in this module; callers apply their own threshold.
+    """
+    img = cv2.imread(str(image_path))
+    if img is None:
+        return None
+    return float(GeminiEngine().detect_watermark(img).confidence)
@@ -0,0 +1,283 @@
+"""Image provenance: identify where an image was made and what watermarks it carries.
+
+Aggregates every locally-readable signal into a single :class:`ProvenanceReport`:
+
+- **C2PA Content Credentials** (issuer, claim generator, digital source type) ->
+  the signing platform (OpenAI, Google, Adobe, Microsoft).
+- **IPTC ``digitalSourceType``** "Made with AI" marker (Meta, X, others).
+- **PNG text / EXIF generation parameters** (Stable Diffusion, ComfyUI, InvokeAI).
+- **SynthID metadata proxy** -- a C2PA companion from a SynthID-using vendor
+  (Google / OpenAI) implies the invisible pixel watermark.
+- **Visible Gemini sparkle** (optional; needs cv2/numpy, no GPU).
+
+Hard limit: a stripped image (re-encoded, screenshotted, social-media upload)
+loses all metadata, and the SynthID *pixel* watermark is not locally decodable
+(proprietary decoder). Absence of signals is therefore reported as ``Unknown``,
+never as "clean". See CLAUDE.md "SynthID detection is metadata-only".
+"""
+
+from __future__ import annotations
+
+import logging
+from dataclasses import dataclass, field
+from typing import TYPE_CHECKING
+
+from remove_ai_watermarks.metadata import (
+    AI_METADATA_KEYS,
+    C2PA_UUID,
+    IPTC_AI_MARKERS,
+    exif_generator,
+    get_ai_metadata,
+)
+from remove_ai_watermarks.noai.c2pa import extract_c2pa_info
+from remove_ai_watermarks.noai.constants import C2PA_AI_TOOLS, C2PA_ISSUERS
+
+if TYPE_CHECKING:
+    from pathlib import Path
+
+log = logging.getLogger(__name__)
+
+# How much of a non-PNG container to binary-scan for the C2PA issuer.
+_SCAN_BYTES = 1024 * 1024
+
+# Visible-sparkle confidence above which the signal is trusted as provenance.
+# Stricter than the removal default (0.25): on the corpus, Gemini-family
+# sparkles score >= 0.56 while non-sparkle images top out at 0.49, so 0.5
+# cleanly separates them and avoids false positives when sparkle is the only
+# signal (e.g. an OpenAI image scored 0.37 -- below threshold, correctly dropped).
+_SPARKLE_THRESHOLD = 0.5
+
+# Issuer (C2PA signer) -> human-readable generating platform. Ordered: when a
+# manifest names several issuers (Microsoft Designer signs as "OpenAI,
+# Microsoft"), the first match wins so the product, not the backend, is named.
+_ISSUER_PLATFORM: tuple[tuple[str, str], ...] = (
+    # Microsoft signs both Designer and Bing Image Creator; Bing now runs its
+    # own MAI-Image model (not DALL-E), so the label stays model-neutral.
+    ("Microsoft", "Microsoft (Bing Image Creator / Designer)"),
+    ("Adobe", "Adobe Firefly"),
+    ("OpenAI", "OpenAI (ChatGPT / gpt-image / DALL-E / Sora)"),
+    ("Google", "Google (Gemini / Imagen)"),
+    ("Stability AI", "Stability AI (Stable Image / DreamStudio)"),
+)
+
+# PNG-text / EXIF keys that indicate a local diffusion pipeline (vs. a hosted
+# platform's C2PA). Subset of AI_METADATA_KEYS; excludes the C2PA/Software keys.
+_LOCAL_GEN_KEYS = frozenset(
+    AI_METADATA_KEYS & {"parameters", "prompt", "negative_prompt", "workflow", "comfyui", "invokeai_metadata", "dream"}
+)
+
+_STRIP_CAVEAT = (
+    "Absence of metadata is not proof the image is clean: C2PA, EXIF, and PNG "
+    "text chunks are stripped by re-encoding, screenshots, or social-media upload."
+)
+_SYNTHID_CAVEAT = (
+    "SynthID is a metadata proxy here; the pixel watermark is not locally "
+    "verifiable (proprietary decoder). Confirm via the Gemini app or openai.com/verify."
+)
+_OPENAI_CAVEAT = (
+    "OpenAI began pairing SynthID with C2PA around 2026-05; OpenAI images from "
+    "before the rollout carry C2PA without SynthID, so the SynthID verdict is 'likely'."
+)
+_IPTC_ONLY_CAVEAT = "The IPTC 'Made with AI' tag flags AI provenance but does not identify the specific platform."
+_INVISIBLE_WM_CAVEAT = (
+    "The open invisible watermark is fragile: it does not survive JPEG re-encoding "
+    "or resizing, so it confirms origin only on a pristine (un-re-encoded) file."
+)
+
+
+@dataclass
+class Signal:
+    """A single provenance signal that was found (or affirmatively absent)."""
+
+    name: str
+    detail: str
+    confidence: str  # "high" | "medium"
+
+
+@dataclass
+class ProvenanceReport:
+    """Aggregated provenance verdict for one image."""
+
+    path: Path
+    is_ai_generated: bool | None  # True / False is never asserted; None = unknown
+    platform: str | None
+    confidence: str  # "high" | "medium" | "none"
+    watermarks: list[str] = field(default_factory=list[str])
+    signals: list[Signal] = field(default_factory=list["Signal"])
+    caveats: list[str] = field(default_factory=list[str])
+
+
+def _issuers_in(data: bytes) -> list[str]:
+    """C2PA issuer names whose signature byte appears in ``data`` (binary scan)."""
+    return sorted({name for sig, name in C2PA_ISSUERS.items() if sig in data})
+
+
+def _ai_tools_in(data: bytes) -> list[str]:
+    """Known C2PA AI-tool / generator names appearing in ``data`` (binary scan).
+
+    PNG has a structured claim_generator; for JPEG/WebP/AVIF/HEIF/JXL the
+    generator lives in a JUMBF/EXIF/XMP blob the PNG parser can't reach, so a
+    byte scan recovers the same attribution (e.g. "Imagen", "DALL-E").
+    """
+    return sorted({name for sig, name in C2PA_AI_TOOLS.items() if sig in data})
+
+
+def _attribute_platform(issuers: list[str]) -> str | None:
+    """Map a set of C2PA issuer names to a human-readable generating platform."""
+    joined = " ".join(issuers)
+    for needle, platform in _ISSUER_PLATFORM:
+        if needle in joined:
+            return platform
+    if issuers:  # e.g. Truepic alone -- a signing authority, not a generator
+        return f"C2PA signer: {', '.join(issuers)} (no known AI generator named)"
+    return None
+
+
+def _visible_sparkle(image_path: Path) -> float | None:
+    """Visible Gemini-sparkle confidence in [0, 1], or None if unavailable.
+
+    Optional: needs cv2/numpy (no GPU). The cv2 work lives in gemini_engine so
+    this module stays dependency-light; returns None if cv2 or the engine
+    assets are missing, or the image can't be read.
+    """
+    try:
+        from remove_ai_watermarks.gemini_engine import detect_sparkle_confidence
+    except Exception as exc:  # cv2/engine assets missing
+        log.debug("visible-sparkle detector unavailable: %s", exc)
+        return None
+    return detect_sparkle_confidence(image_path)
+
+
+def _invisible_watermark(image_path: Path) -> str | None:
+    """Open invisible-watermark scheme name (SD/SDXL/FLUX) or None.
+
+    Optional: needs the imwatermark decoder (extra ``detect``). Returns None if
+    it is not installed or no known watermark decodes.
+    """
+    from remove_ai_watermarks.invisible_watermark import detect_invisible_watermark
+
+    return detect_invisible_watermark(image_path)
+
+
+def identify(image_path: Path, *, check_visible: bool = True, check_invisible: bool = True) -> ProvenanceReport:
+    """Identify an image's origin platform and watermark inventory.
+
+    Args:
+        image_path: Path to the image (PNG, JPEG, WebP, or ISOBMFF container).
+        check_visible: Also run the visible Gemini-sparkle detector (cv2). Set
+            False for a pure-metadata, dependency-light scan.
+        check_invisible: Also decode open invisible watermarks (SD/SDXL/FLUX) via
+            the optional imwatermark library. No-op when it is not installed.
+
+    Returns:
+        A :class:`ProvenanceReport`. ``is_ai_generated`` is True when any AI
+        signal is found and None (unknown) when none is -- it is never asserted
+        False, because stripped metadata leaves no local proof of a clean origin.
+    """
+    info = extract_c2pa_info(image_path)  # PNG-structured; {} for other formats
+    meta = get_ai_metadata(image_path)  # PNG text + EXIF + C2PA fields + synthid
+
+    # First MB covers C2PA (PNG caBX, JPEG APP11, AVIF/HEIF/JXL uuid box) and
+    # IPTC markers for the non-PNG path where extract_c2pa_info returns {}.
+    with open(image_path, "rb") as f:
+        head = f.read(_SCAN_BYTES)
+
+    signals: list[Signal] = []
+    watermarks: list[str] = []
+    caveats: list[str] = []
+
+    # ── C2PA Content Credentials ────────────────────────────────────
+    has_c2pa = bool(info) or b"c2pa" in head.lower() or C2PA_UUID in head
+    issuers = [info["issuer"]] if info.get("issuer") else _issuers_in(head)
+    platform = _attribute_platform(issuers) if has_c2pa else None
+    c2pa_is_ai = "trainedAlgorithmicMedia" in info.get("source_type", "") or any(
+        m in head for m in (b"trainedAlgorithmicMedia", b"compositeWithTrainedAlgorithmicMedia")
+    )
+    # Generator: structured for PNG, binary-scanned for other containers.
+    generator = info.get("claim_generator") or (", ".join(tools) if (tools := _ai_tools_in(head)) else None)
+    if has_c2pa:
+        detail = ", ".join(filter(None, [", ".join(issuers), generator, info.get("source_type")]))
+        signals.append(Signal("c2pa", detail or "C2PA manifest present", "high"))
+        watermarks.append(f"C2PA Content Credentials ({', '.join(issuers) or 'unknown signer'})")
+
+    # ── SynthID metadata proxy ──────────────────────────────────────
+    # get_ai_metadata already sets synthid_watermark for both PNG (caBX parser)
+    # and non-PNG (its own synthid_source fallback), so no extra scan is needed.
+    synthid = meta.get("synthid_watermark")
+    if synthid:
+        watermarks.append(f"SynthID pixel watermark ({synthid})")
+        caveats.append(_SYNTHID_CAVEAT)
+        if "OpenAI" in (" ".join(issuers) + synthid):
+            caveats.append(_OPENAI_CAVEAT)
+
+    # ── IPTC "Made with AI" (Meta etc.), only meaningful without C2PA ─
+    iptc = any(m in head for m in IPTC_AI_MARKERS)
+    if iptc and not has_c2pa:
+        signals.append(Signal("iptc", "digitalSourceType (Made with AI)", "high"))
+        watermarks.append("IPTC digitalSourceType (Made with AI)")
+        caveats.append(_IPTC_ONLY_CAVEAT)
+        if platform is None:
+            platform = "Made-with-AI tag (e.g. Meta AI); platform not specified"
+
+    # ── Local diffusion parameters (Stable Diffusion / ComfyUI) ──────
+    local_keys = sorted(k for k in meta if k.lower() in _LOCAL_GEN_KEYS)
+    if local_keys:
+        signals.append(Signal("gen_params", f"embedded keys: {', '.join(local_keys)}", "high"))
+        watermarks.append("Embedded generation parameters (Stable Diffusion / ComfyUI)")
+        if platform is None:
+            platform = "Stable Diffusion / local pipeline (Automatic1111, ComfyUI, InvokeAI)"
+
+    # ── EXIF Software / XMP CreatorTool generator (cross-format) ─────
+    # Catches a generator tag (incl. inside AVIF/HEIF/JXL) when there is no C2PA.
+    if generator_tag := exif_generator(image_path):
+        signals.append(Signal("exif_generator", f"EXIF/XMP generator: {generator_tag}", "high"))
+        watermarks.append(f"Embedded generator tag: {generator_tag}")
+        if platform is None:
+            platform = f"{generator_tag} (EXIF/XMP generator tag)"
+
+    # ── Open invisible watermark (SD / SDXL / FLUX, dwtDct) ──────────
+    # Public decoder, no key -- a definitive embedded signal on pristine files.
+    if check_invisible and (scheme := _invisible_watermark(image_path)) is not None:
+        signals.append(Signal("invisible_watermark", scheme, "high"))
+        watermarks.append(f"Open invisible watermark: {scheme}")
+        caveats.append(_INVISIBLE_WM_CAVEAT)
+        if platform is None:
+            platform = f"{scheme} (open DWT-DCT watermark)"
+
+    # ── Verdict so far (metadata + embedded watermark) ──────────────
+    invisible_wm = any(s.name == "invisible_watermark" for s in signals)
+    exif_gen = any(s.name == "exif_generator" for s in signals)
+    ai_from_metadata = bool((has_c2pa and (c2pa_is_ai or synthid)) or iptc or local_keys or invisible_wm or exif_gen)
+
+    # ── Visible Gemini sparkle (fallback for stripped-metadata case) ─
+    if check_visible and (conf := _visible_sparkle(image_path)) is not None and conf >= _SPARKLE_THRESHOLD:
+        signals.append(Signal("visible_sparkle", f"NCC confidence {conf:.2f}", "medium"))
+        watermarks.append(f"Visible Gemini sparkle (confidence {conf:.2f})")
+        if platform is None:
+            platform = "Google Gemini family (visible sparkle detected)"
+
+    visible_only = any(s.name == "visible_sparkle" for s in signals) and not ai_from_metadata
+
+    if ai_from_metadata:
+        is_ai: bool | None = True
+        confidence = "high"
+    elif visible_only:
+        is_ai = True
+        confidence = "medium"
+    else:
+        is_ai = None
+        confidence = "none"
+
+    caveats.append(_STRIP_CAVEAT)
+    # De-duplicate while preserving order.
+    caveats = list(dict.fromkeys(caveats))
+
+    return ProvenanceReport(
+        path=image_path,
+        is_ai_generated=is_ai,
+        platform=platform,
+        confidence=confidence,
+        watermarks=watermarks,
+        signals=signals,
+        caveats=caveats,
+    )
@@ -0,0 +1,108 @@
+"""Detect open invisible watermarks embedded by the ``invisible-watermark``
+(imwatermark) library -- used by Stable Diffusion, SDXL, and FLUX.
+
+Unlike SynthID (proprietary, no local decoder), these are DWT-DCT watermarks
+with a PUBLIC decoder and no secret key, so a fresh, un-re-encoded output can be
+identified locally. The known fixed patterns were verified against upstream
+source:
+
+- **Stable Diffusion XL** -- diffusers ``StableDiffusionXLWatermarker``
+  ``WATERMARK_MESSAGE`` (48-bit).
+- **FLUX.2** -- ``black-forest-labs/flux2`` ``src/flux2/watermark.py`` (48-bit).
+- **Stable Diffusion 1.x / 2.x** -- the library's default ``"StableDiffusionV1"``
+  string (136-bit).
+
+The watermark is fragile: it does NOT survive JPEG re-encoding or resizing
+(verified -- gone after JPEG q90), so detection works only on pristine PNG
+originals. Absence is never proof. Requires the optional ``invisible-watermark``
+package (extra: ``detect``); ``detect_invisible_watermark`` returns None when it
+is not installed.
+"""
+
+# imwatermark ships no type stubs (like cv2); its decoder returns are Unknown.
+# Relax the untyped-library diagnostics for this thin wrapper module only.
+# pyright: reportMissingTypeStubs=false, reportUnknownMemberType=false, reportUnknownVariableType=false, reportUnknownArgumentType=false
+
+from __future__ import annotations
+
+import logging
+from typing import TYPE_CHECKING, cast
+
+if TYPE_CHECKING:
+    from pathlib import Path
+
+log = logging.getLogger(__name__)
+
+# Known 48-bit ``bits`` watermarks (dwtDct, no key), name -> message integer.
+_BITS_48: dict[str, int] = {
+    "Stable Diffusion XL": 0b101100111110110010010000011110111011000110011110,
+    "FLUX.2 (Black Forest Labs)": 0b001010101111111010000111100111001111010100101110,
+}
+# The invisible-watermark default string watermark (SD 1.x / 2.x).
+_SD1_STRING = b"StableDiffusionV1"
+
+# Decoded bits/bytes never match a 48-bit pattern by chance: random decode lands
+# near 24/48, an exact embed at 48/48 (measured). 44 (<=4 bit errors) is a safe
+# floor that tolerates light perturbation without risking a false positive.
+_MATCH_48 = 44
+_MATCH_SD1_FRAC = 0.92  # fraction of the 136 string bits that must match
+
+
+def is_available() -> bool:
+    """True if the optional imwatermark decoder is installed."""
+    import importlib.util
+
+    return importlib.util.find_spec("imwatermark") is not None
+
+
+def _bits_match(value: int, ref: int, width: int = 48) -> int:
+    """Number of matching bits between two ``width``-bit integers."""
+    return width - bin(value ^ ref).count("1")
+
+
+def _bytes_match_frac(a: bytes, b: bytes) -> float:
+    """Fraction of matching bits between two equal-length byte strings."""
+    if len(a) != len(b) or not a:
+        return 0.0
+    diff = sum(bin(x ^ y).count("1") for x, y in zip(a, b, strict=True))
+    return 1.0 - diff / (8 * len(b))
+
+
+def detect_invisible_watermark(image_path: Path) -> str | None:
+    """Return the embedding scheme name if a known open watermark is decoded.
+
+    Returns e.g. ``"Stable Diffusion XL"`` / ``"FLUX.2 (Black Forest Labs)"`` /
+    ``"Stable Diffusion 1.x / 2.x"``, or None if none matches, the decoder is
+    unavailable, or the image can't be read. Meaningful only on pristine
+    (un-re-encoded) images.
+    """
+    if not is_available():
+        return None
+    import cv2
+    from imwatermark import WatermarkDecoder
+
+    img = cv2.imread(str(image_path))
+    if img is None:
+        return None
+
+    # 48-bit fixed-message watermarks (SDXL, FLUX.2).
+    try:
+        bits = WatermarkDecoder("bits", 48).decode(img, "dwtDct")
+        value = 0
+        for bit in bits:
+            value = (value << 1) | (1 if bit else 0)
+        for name, ref in _BITS_48.items():
+            if _bits_match(value, ref) >= _MATCH_48:
+                return name
+    except Exception as exc:  # decode can fail on tiny images
+        log.debug("48-bit watermark decode failed for %s: %s", image_path, exc)
+
+    # 136-bit default string watermark (SD 1.x / 2.x).
+    try:
+        raw = cast("bytes", WatermarkDecoder("bytes", 8 * len(_SD1_STRING)).decode(img, "dwtDct"))
+        if _bytes_match_frac(raw, _SD1_STRING) >= _MATCH_SD1_FRAC:
+            return "Stable Diffusion 1.x / 2.x"
+    except Exception as exc:
+        log.debug("string watermark decode failed for %s: %s", image_path, exc)
+
+    return None
@@ -142,80 +142,99 @@ def has_ai_metadata(image_path: Path) -> bool:
    return any(marker in data for marker in IPTC_AI_MARKERS)


-def _scan_png_c2pa_chunk(image_path: Path) -> dict[str, str]:
-    """Extract a human-readable summary of the C2PA manifest in a PNG file.
+def synthid_source(image_path: Path) -> str | None:
+    """Return the vendor name(s) if the image carries a SynthID pixel watermark.

-    PIL does not expose the caBX JUMBF box via ``img.info``, so we delegate
-    chunk extraction to the existing ``extract_c2pa_chunk`` helper and pull
-    key fields from the JUMBF payload without a full CBOR parser.
+    This is a *metadata-based* proxy: Google (Imagen/Gemini) and OpenAI
+    (ChatGPT/DALL-E/gpt-image) embed an invisible SynthID watermark alongside
+    a C2PA manifest, so a C2PA manifest signed by one of them on AI-generated
+    content implies SynthID in the pixels. Adobe Firefly / Microsoft Designer
+    sign C2PA but do not use SynthID, so they return None.
+
+    The verdict is reliable only while the C2PA manifest is intact -- absence
+    is not proof, because C2PA can be stripped while the pixel watermark
+    survives, and the pixel watermark itself is not locally detectable
+    (proprietary decoder).
+
+    Args:
+        image_path: Path to the image (PNG, JPEG, WebP, or ISOBMFF container).
+
+    Returns:
+        Comma-joined vendor name(s) (e.g. ``"OpenAI"``) or None.
+    """
+    from remove_ai_watermarks.noai.c2pa import extract_c2pa_info, synthid_vendors_in
+
+    # PNG: the caBX chunk parser gives a clean, structured issuer.
+    vendors = extract_c2pa_info(image_path).get("synthid_vendors")
+    if vendors:
+        return ", ".join(vendors)
+
+    # Non-PNG containers (JPEG APP11, WebP, AVIF/HEIF/JXL uuid box) keep the
+    # C2PA manifest where the PNG parser can't reach it. Binary-scan for the
+    # same signal: a C2PA manifest from a SynthID-using issuer on AI content.
+    with open(image_path, "rb") as f:
+        data = f.read(1024 * 1024)
+    has_c2pa = b"c2pa" in data.lower() or C2PA_UUID in data
+    # Matches both "trainedAlgorithmicMedia" and "compositeWithTrainedAlgorithmicMedia".
+    ai_source = b"trainedAlgorithmicMedia" in data or b"TrainedAlgorithmicMedia" in data
+    if not (has_c2pa and ai_source):
+        return None
+    matched = synthid_vendors_in(data)
+    return ", ".join(matched) if matched else None
+
+
+def exif_generator(image_path: Path) -> str | None:
+    """Return an AI-generator name from the EXIF ``Software`` / XMP ``CreatorTool``
+    field, if it matches a known generator (see ``AI_GENERATOR_TOKENS``), else None.
+
+    Cross-format: EXIF is read via PIL + piexif for any container PIL can open
+    (JPEG/WebP/AVIF/PNG); an XMP ``CreatorTool`` raw-byte scan additionally covers
+    HEIF/JPEG-XL that PIL can't open without plugins. Only AI tokens match, so
+    ordinary editors (plain "Adobe Photoshop", "GIMP") are not flagged.
    """
    import re

-    from remove_ai_watermarks.noai.c2pa import extract_c2pa_chunk
+    from remove_ai_watermarks.noai.constants import AI_GENERATOR_TOKENS

-    raw = extract_c2pa_chunk(image_path)
-    if raw is None:
-        return {}
+    candidates: list[str] = []

-    # extract_c2pa_chunk returns chunk_header (8 bytes) + data + crc (4 bytes).
-    payload = raw[8:-4]
-    result: dict[str, str] = {"c2pa_manifest": f"C2PA manifest ({len(payload)} bytes)"}
+    # EXIF Software / Artist / ImageDescription (0th IFD) via PIL exif bytes.
+    try:
+        import piexif
+        from PIL import Image

-    def _cbor_text_after(key: bytes) -> str | None:
-        """Return the CBOR text-string immediately following ``key``.
+        with Image.open(image_path) as img:
+            exif_bytes = img.info.get("exif")
+        if exif_bytes:
+            tags = piexif.load(exif_bytes).get("0th", {})
+            # Make catches camera-style tags AI tools reuse (Ideogram writes
+            # Make="Ideogram AI"); real cameras put "Apple"/"Canon" there, which
+            # carry no AI token, so this stays low-false-positive.
+            for tag in (
+                piexif.ImageIFD.Software,
+                piexif.ImageIFD.Make,
+                piexif.ImageIFD.Artist,
+                piexif.ImageIFD.ImageDescription,
+            ):
+                value = tags.get(tag)
+                if isinstance(value, bytes):
+                    candidates.append(value.decode("latin1", "replace"))
+    except Exception as exc:  # unopenable format / malformed EXIF
+        logger.debug("EXIF generator read failed for %s: %s", image_path, exc)

-        Handles CBOR major-type 3 length prefixes: direct (0x60-0x77),
-        1-byte (0x78 NN), and 2-byte (0x79 NN NN).
-        """
-        idx = payload.find(key)
-        if idx < 0:
-            return None
-        p = idx + len(key)
-        if p >= len(payload):
-            return None
-        head = payload[p]
-        if 0x60 <= head <= 0x77:
-            length, start = head - 0x60, p + 1
-        elif head == 0x78 and p + 1 < len(payload):
-            length, start = payload[p + 1], p + 2
-        elif head == 0x79 and p + 2 < len(payload):
-            length, start = (payload[p + 1] << 8) | payload[p + 2], p + 3
-        else:
-            return None
-        raw_str = payload[start : start + length]
-        try:
-            return raw_str.decode("utf-8")
-        except UnicodeDecodeError:
-            return raw_str.decode("latin1", errors="replace")
+    # XMP CreatorTool: text, container-agnostic (covers HEIF/JXL via raw scan).
+    try:
+        with open(image_path, "rb") as f:
+            head = f.read(1024 * 1024)
+        for match in re.finditer(rb"CreatorTool[>\"'=\s]{1,4}([^<\"']{1,80})", head):
+            candidates.append(match.group(1).decode("latin1", "replace"))
+    except Exception as exc:
+        logger.debug("XMP CreatorTool scan failed for %s: %s", image_path, exc)

-    if generator := _cbor_text_after(b"name"):
-        result["claim_generator"] = generator
-
-    if spec := _cbor_text_after(b"specVersion"):
-        result["c2pa_spec"] = spec
-
-    dst_match = re.search(
-        rb"(http://cv\.iptc\.org/newscodes/digitalsourcetype/[A-Za-z0-9_-]+)",
-        payload,
-    )
-    if dst_match:
-        result["digital_source_type"] = dst_match.group(1).decode("latin1")
-
-    actions = sorted(
-        {m.decode("latin1") for m in re.findall(rb"c2pa\.(created|converted|edited|opened|placed)", payload)}
-    )
-    if actions:
-        result["c2pa_actions"] = ", ".join(actions)
-
-    # Scan cert DN printable strings for the signer org name.
-    signer_match = re.search(
-        rb"([A-Za-z][A-Za-z0-9 .,&'()\-]{2,48}OpenAI[A-Za-z0-9 .,&'()\-]{0,48})",
-        payload,
-    )
-    if signer_match:
-        result["signer"] = signer_match.group(1).decode("latin1").strip()
-
-    return result
+    for value in candidates:
+        if any(token in value.lower() for token in AI_GENERATOR_TOKENS):
+            return value.strip()
+    return None


 def get_ai_metadata(image_path: Path) -> dict[str, str]:
@@ -229,19 +248,44 @@ def get_ai_metadata(image_path: Path) -> dict[str, str]:
    """
    from PIL import Image

+    from remove_ai_watermarks.noai.c2pa import extract_c2pa_info, synthid_verdict
+
    result: dict[str, str] = {}

-    with Image.open(image_path) as img:
-        for key, value in img.info.items():
-            if _is_ai_key(key):
-                if isinstance(value, bytes):
-                    result[key] = f"<binary {len(value)} bytes>"
-                elif isinstance(value, str) and len(value) > 200:
-                    result[key] = value[:200] + "…"
-                else:
-                    result[key] = str(value)
+    # PIL may not open AVIF/HEIF/JPEG-XL without optional plugins (and
+    # ultralytics' Image.open patch can raise ModuleNotFoundError); fall through
+    # to the C2PA/binary path on any open failure. See CLAUDE.md.
+    try:
+        with Image.open(image_path) as img:
+            for key, value in img.info.items():
+                if _is_ai_key(key):
+                    if isinstance(value, bytes):
+                        result[key] = f"<binary {len(value)} bytes>"
+                    elif isinstance(value, str) and len(value) > 200:
+                        result[key] = value[:200] + "…"
+                    else:
+                        result[key] = str(value)
+    except Exception as exc:
+        logger.debug("PIL could not open %s for AI-metadata scan: %s", image_path, exc)

-    result.update(_scan_png_c2pa_chunk(image_path))
+    # C2PA manifest fields from the single canonical parser (noai/c2pa.py).
+    c2pa = extract_c2pa_info(image_path)
+    for key in (
+        "c2pa_manifest",
+        "claim_generator",
+        "c2pa_spec",
+        "issuer",
+        "source_type",
+        "actions",
+        "synthid_watermark",
+    ):
+        if key in c2pa:
+            result.setdefault(key, str(c2pa[key]))
+
+    # Non-PNG containers (JPEG/WebP/AVIF): extract_c2pa_info is PNG-only, so
+    # fall back to the format-agnostic source check for the SynthID verdict.
+    if "synthid_watermark" not in result and (vendor := synthid_source(image_path)):
+        result.setdefault("synthid_watermark", synthid_verdict(vendor))
    return result


@@ -29,6 +29,7 @@ from remove_ai_watermarks.noai.constants import (
    C2PA_ISSUERS,
    C2PA_SIGNATURES,
    PNG_SIGNATURE,
+    SYNTHID_C2PA_ISSUERS,
 )


@@ -129,10 +130,56 @@ def extract_c2pa_info(image_path: Path) -> dict[str, Any]:
    return c2pa_info


+def _cbor_text_after(payload: bytes, key: bytes) -> str | None:
+    """Return the CBOR text-string immediately following ``key`` in ``payload``.
+
+    Handles CBOR major-type 3 length prefixes: direct (0x60-0x77), 1-byte
+    (0x78 NN), and 2-byte (0x79 NN NN). This reads the actual encoded value, so
+    it avoids the byte-grabbing artifacts a loose regex produces (e.g. the
+    leading length byte showing up as ``fGPT-4o``).
+    """
+    idx = payload.find(key)
+    if idx < 0:
+        return None
+    p = idx + len(key)
+    if p >= len(payload):
+        return None
+    head = payload[p]
+    if 0x60 <= head <= 0x77:
+        length, start = head - 0x60, p + 1
+    elif head == 0x78 and p + 1 < len(payload):
+        length, start = payload[p + 1], p + 2
+    elif head == 0x79 and p + 2 < len(payload):
+        length, start = (payload[p + 1] << 8) | payload[p + 2], p + 3
+    else:
+        return None
+    raw_str = payload[start : start + length]
+    try:
+        return raw_str.decode("utf-8")
+    except UnicodeDecodeError:
+        return raw_str.decode("latin1", errors="replace")
+
+
+def synthid_verdict(vendors: str) -> str:
+    """Human-readable SynthID-source verdict, shared by all callers."""
+    return f"likely present ({vendors} embeds SynthID with C2PA)"
+
+
+def synthid_vendors_in(buffer: bytes) -> list[str]:
+    """Return SynthID-using C2PA issuer names whose signature appears in ``buffer``.
+
+    Shared by the PNG caBX parser and the format-agnostic binary scan so both
+    apply the same SYNTHID_C2PA_ISSUERS rule against their respective bytes.
+    """
+    return sorted({name for sig, name in C2PA_ISSUERS.items() if sig in buffer and sig in SYNTHID_C2PA_ISSUERS})
+
+
 def _parse_c2pa_chunk(chunk_data: bytes, c2pa_info: dict[str, Any]) -> None:
    """Parse C2PA chunk data and populate info dictionary."""
+    c2pa_info["c2pa_manifest"] = f"C2PA manifest ({len(chunk_data)} bytes)"
+
    # Find issuers
-    issuers = []
+    issuers: list[str] = []
    for sig, name in C2PA_ISSUERS.items():
        if sig in chunk_data:
            issuers.append(name)
@@ -140,44 +187,25 @@ def _parse_c2pa_chunk(chunk_data: bytes, c2pa_info: dict[str, Any]) -> None:
        c2pa_info["issuer"] = ", ".join(set(issuers))

    # Find AI tools
-    ai_tools = []
+    ai_tools: list[str] = []
    for sig, name in C2PA_AI_TOOLS.items():
        if sig in chunk_data:
            ai_tools.append(name)
    if ai_tools:
        c2pa_info["ai_tool"] = ", ".join(set(ai_tools))

-    # Extract software agent (multiple patterns)
-    patterns = [
-        rb"softwareAgent.*?dname([^\x00]+?)(?:q|l|m|n)",
-        rb"software_agent[^\x00]*?([A-Za-z0-9_\-\.]+)",
-        rb"Software[^\x00]*?([A-Za-z0-9_\-\. ]+)",
-    ]
-    for pattern in patterns:
-        match = re.search(pattern, chunk_data, re.DOTALL | re.IGNORECASE)
-        if match:
-            agent = match.group(1).decode("utf-8", errors="ignore").strip()
-            if agent and len(agent) < 100:
-                c2pa_info["software_agent"] = agent
-                break
-
-    # Extract claim generator (multiple patterns)
-    claim_patterns = [
-        rb"claim_generator[^\x00]*?([A-Za-z0-9_\-\.\/\:]+)",
-        rb"claimGenerator[^\x00]*?([A-Za-z0-9_\-\.\/\:]+)",
-        rb"dname([^\x00]{3,50})(?:q|l|m|n|i)",
-    ]
-    for pattern in claim_patterns:
-        match = re.search(pattern, chunk_data, re.DOTALL | re.IGNORECASE)
-        if match:
-            gen_name = match.group(1).decode("utf-8", errors="ignore").strip()
-            # Filter out common false positives
-            if gen_name and len(gen_name) < 100 and not gen_name.startswith(("\\x", "\\\\x")):
-                c2pa_info["claim_generator"] = gen_name
-                break
+    # Claim generator and spec version: read the CBOR text-string values
+    # directly (regex byte-grabbing produced artifacts like ``fGPT-4o``).
+    # Guard with isprintable(): on some manifests (e.g. Microsoft Designer) the
+    # first ``name`` key precedes a binary field (a hash), not the generator
+    # string, which would otherwise surface as control-char garbage.
+    if (generator := _cbor_text_after(chunk_data, b"name")) and generator.isprintable():
+        c2pa_info["claim_generator"] = generator
+    if (spec := _cbor_text_after(chunk_data, b"specVersion")) and spec.isprintable():
+        c2pa_info["c2pa_spec"] = spec

    # Find actions
-    actions = []
+    actions: list[str] = []
    for sig, name in C2PA_ACTIONS.items():
        if sig in chunk_data:
            actions.append(name)
@@ -192,12 +220,23 @@ def _parse_c2pa_chunk(chunk_data: bytes, c2pa_info: dict[str, Any]) -> None:
            c2pa_info["timestamps"] = [t.decode("utf-8") for t in timestamp_matches[:3]]

    # Find digital source type
+    ai_source = False
    if b"trainedAlgorithmicMedia" in chunk_data:
        c2pa_info["source_type"] = "trainedAlgorithmicMedia (AI-generated)"
+        ai_source = True
    elif b"algorithmicMedia" in chunk_data:
        c2pa_info["source_type"] = "algorithmicMedia"
    elif b"compositeWithTrainedAlgorithmicMedia" in chunk_data:
        c2pa_info["source_type"] = "compositeWithTrainedAlgorithmicMedia (AI-enhanced)"
+        ai_source = True
+
+    # SynthID pixel-watermark proxy: a C2PA manifest from a SynthID-using
+    # vendor (Google/OpenAI) on AI-generated content implies an invisible
+    # SynthID watermark in the pixels (see SYNTHID_C2PA_ISSUERS).
+    synthid_vendors = synthid_vendors_in(chunk_data)
+    if synthid_vendors and ai_source:
+        c2pa_info["synthid_vendors"] = synthid_vendors
+        c2pa_info["synthid_watermark"] = synthid_verdict(", ".join(synthid_vendors))


 def extract_c2pa_chunk(image_path: Path) -> bytes | None:
@@ -85,8 +85,32 @@ C2PA_ISSUERS = {
    b"Microsoft": "Microsoft",
    b"OpenAI": "OpenAI",
    b"Truepic": "Truepic",
+    # Stability AI signs C2PA as "Stability AI" (cert org "Stability AI Ltd").
+    # Verified on a live Brand Studio (DreamStudio successor) output, 2026-05-24.
+    b"Stability AI": "Stability AI",
 }

+# C2PA issuers whose signed outputs also carry an invisible SynthID pixel
+# watermark -- a metadata proxy for "SynthID is in the pixels":
+#   - Google (Imagen/Gemini): embeds SynthID, long-standing (DeepMind docs).
+#   - OpenAI (ChatGPT/Codex/API): pairs SynthID with C2PA since ~2026-05-20.
+#     Confirmed by OpenAI's Help Center ("C2PA and SynthID in OpenAI-generated
+#     images", updated 2026-05-21): "Images generated with ChatGPT, Codex, and
+#     our API include both C2PA metadata and SynthID watermarks." OpenAI also
+#     notes a signal may be absent if "the image was created before these
+#     signals were available" -- so OpenAI images from BEFORE the rollout carry
+#     C2PA WITHOUT SynthID (e.g. data/samples/openai-images-2/amur-leopard.png,
+#     C2PA timestamp 2026-04-22). For OpenAI the proxy is therefore "likely",
+#     not certain; the verdict string is hedged accordingly. OpenAI's own oracle
+#     is openai.com/verify (Google's is the Gemini app "Verify with SynthID").
+# The issuer byte ("OpenAI"/"Google") is verified locally against data/samples;
+# the SynthID pairing is documented behavior (Google: DeepMind; OpenAI: above).
+# Adobe Firefly and Microsoft Designer sign C2PA but do NOT use SynthID, so a
+# C2PA manifest alone is not a SynthID signal -- the issuer is. The pixel
+# watermark is not locally detectable (proprietary decoder); the C2PA companion
+# is the proxy, and only while the manifest is intact.
+SYNTHID_C2PA_ISSUERS = frozenset({b"Google", b"OpenAI"})
+
 # C2PA known AI tools
 C2PA_AI_TOOLS = {
    b"GPT-4o": "GPT-4o",
@@ -98,6 +122,32 @@ C2PA_AI_TOOLS = {
    b"Firefly": "Firefly",
 }

+# Lowercased substrings that mark an AI generator when found in an EXIF
+# ``Software`` / XMP ``CreatorTool`` value. Conservative on purpose: plain
+# editors like "Adobe Photoshop" or "GIMP" must NOT match (no AI token), so only
+# generator names land here. Add new generators here, not inline.
+AI_GENERATOR_TOKENS: frozenset[str] = frozenset(
+    {
+        "firefly",
+        "dall-e",
+        "dalle",
+        "midjourney",
+        "stable diffusion",
+        "stable-diffusion",
+        "stablediffusion",
+        "comfyui",
+        "automatic1111",
+        "invokeai",
+        "imagen",
+        "gpt-image",
+        "nightcafe",
+        "ideogram",
+        "leonardo",
+        "flux",
+        "dreamstudio",
+    }
+)
+
 # C2PA action types
 C2PA_ACTIONS = {
    b"c2pa.created": "created",
@@ -106,6 +156,8 @@ C2PA_ACTIONS = {
    b"c2pa.filtered": "filtered",
    b"c2pa.cropped": "cropped",
    b"c2pa.resized": "resized",
+    b"c2pa.opened": "opened",
+    b"c2pa.placed": "placed",
 }

 # PNG signature
@@ -2,6 +2,7 @@

 from __future__ import annotations

+import json
 from typing import TYPE_CHECKING
 from unittest.mock import MagicMock, patch

@@ -243,6 +244,36 @@ class TestMetadataCommand:
        assert "stripped" in result.output


+class TestIdentifyCommand:
+    """Tests for the 'identify' subcommand."""
+
+    def test_identify_help(self, runner):
+        result = runner.invoke(main, ["identify", "--help"])
+        assert result.exit_code == 0
+
+    def test_identify_clean_png(self, runner, tmp_clean_png):
+        result = runner.invoke(main, ["identify", str(tmp_clean_png), "--no-visible"])
+        assert result.exit_code == 0
+        assert "unknown" in result.output
+
+    def test_identify_ai_png_reports_platform(self, runner, tmp_png_with_ai_metadata):
+        result = runner.invoke(main, ["identify", str(tmp_png_with_ai_metadata), "--no-visible"])
+        assert result.exit_code == 0
+        assert "AI-generated" in result.output
+        assert "Stable Diffusion" in result.output
+
+    def test_identify_json_is_valid(self, runner, tmp_png_with_ai_metadata):
+        result = runner.invoke(main, ["identify", str(tmp_png_with_ai_metadata), "--no-visible", "--json"])
+        assert result.exit_code == 0
+        payload = json.loads(result.output)
+        assert payload["is_ai_generated"] is True
+        assert payload["confidence"] == "high"
+
+    def test_identify_nonexistent_file(self, runner):
+        result = runner.invoke(main, ["identify", "/nonexistent/file.png"])
+        assert result.exit_code != 0
+
+
 class TestBatchCommand:
    """Tests for the 'batch' subcommand."""

@@ -12,6 +12,7 @@ from remove_ai_watermarks.gemini_engine import (
    WatermarkPosition,
    WatermarkSize,
    _calculate_alpha_map,
+    detect_sparkle_confidence,
    get_watermark_config,
    get_watermark_size,
 )
@@ -174,6 +175,21 @@ class TestDetection:
        assert isinstance(result.gradient_score, float)


+class TestDetectSparkleConfidence:
+    """File-level entry point used by identify.py."""
+
+    def test_returns_float_in_range_for_real_image(self, tmp_image_path):
+        conf = detect_sparkle_confidence(tmp_image_path)
+        assert conf is not None
+        assert 0.0 <= conf <= 1.0
+
+    def test_returns_none_for_unreadable_file(self, tmp_path):
+        # cv2.imread returns None for a non-image; the helper must not raise.
+        bogus = tmp_path / "not_an_image.png"
+        bogus.write_bytes(b"this is not a PNG")
+        assert detect_sparkle_confidence(bogus) is None
+
+
 # ── Inpainting ──────────────────────────────────────────────────────


@@ -0,0 +1,319 @@
+"""Tests for the provenance identifier (identify.py).
+
+Pure attribution logic is unit-tested directly; end-to-end verdicts assert
+against the real committed C2PA / IPTC fixtures in data/samples/.
+"""
+
+from __future__ import annotations
+
+import json
+from dataclasses import asdict
+from pathlib import Path
+from unittest.mock import patch
+
+import pytest
+
+from remove_ai_watermarks.identify import (
+    ProvenanceReport,
+    _ai_tools_in,
+    _attribute_platform,
+    _issuers_in,
+    identify,
+)
+
+# Where the lazy import inside identify._visible_sparkle resolves the detector.
+_SPARKLE_TARGET = "remove_ai_watermarks.gemini_engine.detect_sparkle_confidence"
+
+SAMPLES_DIR = Path(__file__).resolve().parent.parent / "data" / "samples"
+
+
+# ── Pure attribution logic (no file IO) ─────────────────────────────
+
+
+class TestAttributePlatform:
+    def test_openai(self):
+        assert "OpenAI" in (_attribute_platform(["OpenAI"]) or "")
+
+    def test_designer_wins_over_openai_backend(self):
+        # Microsoft Designer signs as "OpenAI, Microsoft"; name the product.
+        platform = _attribute_platform(["OpenAI", "Microsoft"])
+        assert platform
+        assert "Designer" in platform
+
+    def test_adobe(self):
+        assert _attribute_platform(["Adobe"]) == "Adobe Firefly"
+
+    def test_google(self):
+        assert "Google" in (_attribute_platform(["Google LLC"]) or "")
+
+    def test_truepic_is_signer_not_generator(self):
+        platform = _attribute_platform(["Truepic"])
+        assert platform
+        assert "signer" in platform.lower()
+
+    def test_microsoft_label_is_model_neutral(self):
+        # Bing now runs MAI-Image, not DALL-E; the label must not claim DALL-E.
+        platform = _attribute_platform(["Microsoft"])
+        assert platform
+        assert "DALL-E" not in platform
+
+    def test_stability(self):
+        platform = _attribute_platform(["Stability AI"])
+        assert platform
+        assert "Stability AI" in platform
+
+    def test_empty_is_none(self):
+        assert _attribute_platform([]) is None
+
+
+class TestIssuersIn:
+    def test_finds_openai(self):
+        assert _issuers_in(b"...OpenAI...trainedAlgorithmicMedia") == ["OpenAI"]
+
+    def test_finds_multiple_sorted(self):
+        assert _issuers_in(b"Microsoft and OpenAI") == ["Microsoft", "OpenAI"]
+
+    def test_none_present(self):
+        assert _issuers_in(b"just some bytes") == []
+
+
+class TestAiToolsIn:
+    def test_finds_generator(self):
+        assert _ai_tools_in(b"...claim_generator Imagen 3...") == ["Imagen"]
+
+    def test_none_present(self):
+        assert _ai_tools_in(b"a regular photo, no tools") == []
+
+
+class TestIdentifyNonPng:
+    """Non-PNG containers (JPEG/WebP/AVIF) carry C2PA where the caBX parser can't
+    reach; identify recovers issuer + generator via the binary scan. Synthetic
+    byte blobs mirror tests/test_metadata.py::TestSynthIDSourceNonPng.
+    """
+
+    def _c2pa_jpeg(self, tmp_path: Path, blob: bytes) -> Path:
+        path = tmp_path / "img.jpg"
+        path.write_bytes(b"\xff\xd8\xff\xe1jumbc2pa" + blob + b"\xff\xd9")
+        return path
+
+    def test_google_imagen_jpeg(self, tmp_path: Path):
+        path = self._c2pa_jpeg(tmp_path, b"Google Imagen ... trainedAlgorithmicMedia")
+        r = identify(path, check_visible=False)
+        assert r.is_ai_generated is True
+        assert r.platform is not None
+        assert "Google" in r.platform
+        # Generator recovered from the non-PNG blob shows up in the c2pa signal.
+        c2pa_signal = next(s for s in r.signals if s.name == "c2pa")
+        assert "Imagen" in c2pa_signal.detail
+
+    def test_openai_jpeg_has_synthid(self, tmp_path: Path):
+        path = self._c2pa_jpeg(tmp_path, b"OpenAI DALL-E ... trainedAlgorithmicMedia")
+        r = identify(path, check_visible=False)
+        assert any("SynthID" in w for w in r.watermarks)
+
+    def test_stability_ai_issuer_attributed_no_synthid(self, tmp_path: Path):
+        path = self._c2pa_jpeg(tmp_path, b"Stability AI ... trainedAlgorithmicMedia")
+        r = identify(path, check_visible=False)
+        assert r.is_ai_generated is True
+        assert r.platform is not None
+        assert "Stability AI" in r.platform
+        assert not any("SynthID" in w for w in r.watermarks)  # Stability does not use SynthID
+
+    def test_c2pa_without_ai_marker_is_unknown(self, tmp_path: Path):
+        # Adobe signs C2PA on plain Photoshop edits too. Without an AI digital-
+        # source marker, the honest verdict is unknown -- the C2PA watermark is
+        # still listed, but is_ai_generated is not asserted True.
+        path = self._c2pa_jpeg(tmp_path, b"Adobe ... no ai marker here")
+        r = identify(path, check_visible=False)
+        assert r.is_ai_generated is None
+        assert any("C2PA" in w for w in r.watermarks)
+        assert not any("SynthID" in w for w in r.watermarks)
+
+
+# ── End-to-end verdicts on real fixtures ────────────────────────────
+
+
+@pytest.mark.skipif(not SAMPLES_DIR.exists(), reason="data/samples not present")
+class TestIdentifyRealSamples:
+    def test_openai_chatgpt(self):
+        r = identify(SAMPLES_DIR / "chatgpt-1.png", check_visible=False)
+        assert r.is_ai_generated is True
+        assert r.confidence == "high"
+        assert r.platform
+        assert "OpenAI" in r.platform
+        assert any("C2PA" in w for w in r.watermarks)
+        assert any("SynthID" in w for w in r.watermarks)
+
+    def test_adobe_firefly_has_no_synthid(self):
+        r = identify(SAMPLES_DIR / "firefly-1.png", check_visible=False)
+        assert r.is_ai_generated is True
+        assert r.platform == "Adobe Firefly"
+        assert not any("SynthID" in w for w in r.watermarks)
+
+    def test_iptc_made_with_ai(self):
+        # mj-1.png carries the IPTC digitalSourceType "Made with AI" marker.
+        r = identify(SAMPLES_DIR / "mj-1.png", check_visible=False)
+        assert r.is_ai_generated is True
+        assert any("IPTC" in w for w in r.watermarks)
+
+    def test_clean_photo_is_unknown_not_clean(self):
+        r = identify(SAMPLES_DIR / "not-ai-1.jpeg", check_visible=False)
+        assert r.is_ai_generated is None  # never asserted False
+        assert r.platform is None
+        assert r.confidence == "none"
+        assert r.watermarks == []
+
+    def test_strip_caveat_always_present(self):
+        r = identify(SAMPLES_DIR / "not-ai-1.jpeg", check_visible=False)
+        assert any("not proof" in c for c in r.caveats)
+
+    def test_returns_report_dataclass(self):
+        assert isinstance(identify(SAMPLES_DIR / "firefly-1.png", check_visible=False), ProvenanceReport)
+
+
+# ── Local diffusion parameters (Stable Diffusion / ComfyUI) ─────────
+
+
+class TestIdentifyLocalParams:
+    """A PNG carrying SD-style generation params is attributed to a local pipeline."""
+
+    def test_sd_params_attributed_to_local_pipeline(self, tmp_png_with_ai_metadata: Path):
+        r = identify(tmp_png_with_ai_metadata, check_visible=False)
+        assert r.is_ai_generated is True
+        assert r.confidence == "high"
+        assert r.platform is not None
+        assert "Stable Diffusion" in r.platform
+        assert any("generation parameters" in w for w in r.watermarks)
+
+    def test_gen_params_signal_lists_keys(self, tmp_png_with_ai_metadata: Path):
+        r = identify(tmp_png_with_ai_metadata, check_visible=False)
+        signal = next(s for s in r.signals if s.name == "gen_params")
+        assert "parameters" in signal.detail
+        assert signal.confidence == "high"
+
+    def test_clean_png_is_unknown(self, tmp_clean_png: Path):
+        r = identify(tmp_clean_png, check_visible=False)
+        assert r.is_ai_generated is None
+        assert r.platform is None
+        assert r.confidence == "none"
+        assert r.signals == []
+
+
+# ── Visible-sparkle fallback (mocked detector) ──────────────────────
+
+
+class TestIdentifyVisibleSparkle:
+    """The visible-sparkle signal gates on the corpus-tuned threshold (0.5)."""
+
+    def test_above_threshold_promotes_to_medium(self, tmp_clean_png: Path):
+        with patch(_SPARKLE_TARGET, return_value=0.7):
+            r = identify(tmp_clean_png, check_visible=True)
+        assert r.is_ai_generated is True
+        assert r.confidence == "medium"
+        assert r.platform is not None
+        assert "Gemini" in r.platform
+        signal = next(s for s in r.signals if s.name == "visible_sparkle")
+        assert signal.confidence == "medium"
+
+    def test_below_threshold_not_promoted(self, tmp_clean_png: Path):
+        with patch(_SPARKLE_TARGET, return_value=0.4):
+            r = identify(tmp_clean_png, check_visible=True)
+        assert r.is_ai_generated is None
+        assert not any(s.name == "visible_sparkle" for s in r.signals)
+
+    def test_detector_unavailable_does_not_crash(self, tmp_clean_png: Path):
+        with patch(_SPARKLE_TARGET, return_value=None):
+            r = identify(tmp_clean_png, check_visible=True)
+        assert r.is_ai_generated is None
+        assert not any(s.name == "visible_sparkle" for s in r.signals)
+
+    def test_check_visible_false_skips_detector(self, tmp_clean_png: Path):
+        # Even a strong detection is ignored when the caller opts out.
+        with patch(_SPARKLE_TARGET, return_value=0.99) as mock_detect:
+            r = identify(tmp_clean_png, check_visible=False)
+        mock_detect.assert_not_called()
+        assert not any(s.name == "visible_sparkle" for s in r.signals)
+
+    def test_metadata_keeps_high_even_with_sparkle(self, tmp_png_with_ai_metadata: Path):
+        # Metadata verdict (high) is not downgraded by an additional sparkle hit.
+        with patch(_SPARKLE_TARGET, return_value=0.7):
+            r = identify(tmp_png_with_ai_metadata, check_visible=True)
+        assert r.confidence == "high"
+
+
+# ── Caveats and serialization ───────────────────────────────────────
+
+
+@pytest.mark.skipif(not SAMPLES_DIR.exists(), reason="data/samples not present")
+class TestIdentifyCaveats:
+    def test_openai_hedge_caveat_present(self):
+        r = identify(SAMPLES_DIR / "chatgpt-1.png", check_visible=False)
+        assert any("before the rollout" in c for c in r.caveats)
+
+    def test_synthid_proxy_caveat_present(self):
+        r = identify(SAMPLES_DIR / "chatgpt-1.png", check_visible=False)
+        assert any("not locally" in c for c in r.caveats)
+
+    def test_caveats_are_deduplicated(self):
+        r = identify(SAMPLES_DIR / "chatgpt-1.png", check_visible=False)
+        assert len(r.caveats) == len(set(r.caveats))
+
+
+class TestReportSerializable:
+    def test_report_is_json_serializable(self, tmp_png_with_ai_metadata: Path):
+        # The CLI --json path relies on asdict + json.dumps(default=str).
+        report = identify(tmp_png_with_ai_metadata, check_visible=False)
+        dumped = json.dumps(asdict(report), default=str)
+        assert "is_ai_generated" in dumped
+
+
+class TestIdentifyExifGenerator:
+    """An AI generator tag in EXIF/XMP (incl. AVIF) drives attribution."""
+
+    def test_avif_firefly_software_attributed(self, tmp_path: Path):
+        import piexif
+        from PIL import Image
+
+        exif = piexif.dump({"0th": {piexif.ImageIFD.Software: b"Adobe Firefly"}, "Exif": {}, "GPS": {}, "1st": {}})
+        path = tmp_path / "firefly.avif"
+        Image.new("RGB", (64, 64), (90, 80, 70)).save(path, exif=exif)
+        r = identify(path, check_visible=False)
+        assert r.is_ai_generated is True
+        assert r.platform is not None
+        assert "Firefly" in r.platform
+        assert any("generator tag" in w for w in r.watermarks)
+
+
+# ── Open invisible watermark (SD/SDXL/FLUX) integration ─────────────
+
+from remove_ai_watermarks.invisible_watermark import is_available as _wm_available  # noqa: E402
+
+
+@pytest.mark.skipif(not _wm_available(), reason="invisible-watermark not installed")
+class TestIdentifyInvisibleWatermark:
+    def _sdxl_watermarked(self, tmp_path: Path) -> Path:
+        import cv2
+        import numpy as np
+        from imwatermark import WatermarkEncoder
+
+        from remove_ai_watermarks.invisible_watermark import _BITS_48
+
+        bits = [int(b) for b in format(_BITS_48["Stable Diffusion XL"], "048b")]
+        enc = WatermarkEncoder()
+        enc.set_watermark("bits", bits)
+        img = np.random.default_rng(0).integers(0, 255, (512, 512, 3), dtype=np.uint8)
+        path = tmp_path / "sdxl.png"
+        cv2.imwrite(str(path), enc.encode(img, "dwtDct"))
+        return path
+
+    def test_sdxl_watermark_identified(self, tmp_path: Path):
+        r = identify(self._sdxl_watermarked(tmp_path), check_visible=False)
+        assert r.is_ai_generated is True
+        assert r.confidence == "high"
+        assert r.platform is not None
+        assert "Stable Diffusion XL" in r.platform
+        assert any("invisible watermark" in w.lower() for w in r.watermarks)
+
+    def test_check_invisible_false_skips(self, tmp_path: Path):
+        r = identify(self._sdxl_watermarked(tmp_path), check_visible=False, check_invisible=False)
+        assert not any(s.name == "invisible_watermark" for s in r.signals)
@@ -0,0 +1,90 @@
+"""Tests for open invisible-watermark (imwatermark) detection.
+
+Each known scheme is round-tripped: embed its exact upstream pattern with the
+encoder, then assert the detector names it. Skipped entirely if the optional
+``invisible-watermark`` package is not installed.
+"""
+
+from __future__ import annotations
+
+from typing import TYPE_CHECKING
+
+import cv2
+import numpy as np
+import pytest
+
+if TYPE_CHECKING:
+    from pathlib import Path
+
+from remove_ai_watermarks.invisible_watermark import (
+    _BITS_48,
+    _SD1_STRING,
+    _bits_match,
+    _bytes_match_frac,
+    detect_invisible_watermark,
+    is_available,
+)
+
+pytestmark = pytest.mark.skipif(not is_available(), reason="invisible-watermark not installed")
+
+
+def _base_image() -> np.ndarray:
+    # imwatermark needs enough DWT coefficients; use a 512x512 textured image.
+    rng = np.random.default_rng(0)
+    return rng.integers(0, 255, (512, 512, 3), dtype=np.uint8)
+
+
+def _write_bits_watermark(tmp_path: Path, message: int) -> Path:
+    from imwatermark import WatermarkEncoder
+
+    bits = [int(b) for b in format(message, "048b")]
+    enc = WatermarkEncoder()
+    enc.set_watermark("bits", bits)
+    wm = enc.encode(_base_image(), "dwtDct")
+    path = tmp_path / "wm.png"
+    cv2.imwrite(str(path), wm)
+    return path
+
+
+class TestHelpers:
+    def test_bits_match_exact(self):
+        assert _bits_match(0b1010, 0b1010, width=4) == 4
+
+    def test_bits_match_one_off(self):
+        assert _bits_match(0b1010, 0b1011, width=4) == 3
+
+    def test_bytes_match_identical(self):
+        assert _bytes_match_frac(_SD1_STRING, _SD1_STRING) == 1.0
+
+    def test_bytes_match_length_mismatch_is_zero(self):
+        assert _bytes_match_frac(b"abc", b"abcd") == 0.0
+
+
+class TestDetect:
+    def test_detects_sdxl(self, tmp_path: Path):
+        path = _write_bits_watermark(tmp_path, _BITS_48["Stable Diffusion XL"])
+        assert detect_invisible_watermark(path) == "Stable Diffusion XL"
+
+    def test_detects_flux(self, tmp_path: Path):
+        path = _write_bits_watermark(tmp_path, _BITS_48["FLUX.2 (Black Forest Labs)"])
+        assert detect_invisible_watermark(path) == "FLUX.2 (Black Forest Labs)"
+
+    def test_detects_sd1_string(self, tmp_path: Path):
+        from imwatermark import WatermarkEncoder
+
+        enc = WatermarkEncoder()
+        enc.set_watermark("bytes", _SD1_STRING)
+        wm = enc.encode(_base_image(), "dwtDct")
+        path = tmp_path / "sd1.png"
+        cv2.imwrite(str(path), wm)
+        assert detect_invisible_watermark(path) == "Stable Diffusion 1.x / 2.x"
+
+    def test_clean_image_is_none(self, tmp_path: Path):
+        path = tmp_path / "clean.png"
+        cv2.imwrite(str(path), _base_image())
+        assert detect_invisible_watermark(path) is None
+
+    def test_unreadable_file_is_none(self, tmp_path: Path):
+        path = tmp_path / "not_image.png"
+        path.write_bytes(b"not a png")
+        assert detect_invisible_watermark(path) is None
@@ -4,16 +4,23 @@ from __future__ import annotations

 from pathlib import Path

+import piexif
+import pytest
 from PIL import Image
 from PIL.PngImagePlugin import PngInfo

 from remove_ai_watermarks.metadata import (
    _is_ai_key,
+    exif_generator,
    get_ai_metadata,
    has_ai_metadata,
    remove_ai_metadata,
+    synthid_source,
 )

+# Real, committed C2PA sample images used to ground the SynthID-source tests.
+SAMPLES_DIR = Path(__file__).resolve().parent.parent / "data" / "samples"
+
 # ── Key detection ───────────────────────────────────────────────────


@@ -144,6 +151,116 @@ class TestGetAiMetadata:
        meta = get_ai_metadata(tmp_clean_png)
        assert meta == {}

+    def test_long_value_is_truncated(self, tmp_path: Path):
+        img = Image.new("RGB", (32, 32))
+        pnginfo = PngInfo()
+        pnginfo.add_text("parameters", "x" * 300)
+        path = tmp_path / "long.png"
+        img.save(path, pnginfo=pnginfo)
+        meta = get_ai_metadata(path)
+        assert meta["parameters"].endswith("…")
+        assert len(meta["parameters"]) <= 205
+
+    def test_unopenable_file_does_not_raise(self, tmp_path: Path):
+        # PIL can't open HEIC without pillow-heif; get_ai_metadata must fall
+        # through to the binary scan, not propagate UnidentifiedImageError.
+        path = tmp_path / "iphone.heic"
+        path.write_bytes(b"\x00\x00\x00\x18ftypheic" + b"\x00" * 64)
+        assert get_ai_metadata(path) == {}
+
+
+@pytest.mark.skipif(not SAMPLES_DIR.exists(), reason="data/samples not present")
+class TestGetAiMetadataRealSample:
+    """get_ai_metadata surfaces the consolidated C2PA fields on real images."""
+
+    def test_openai_sample_fields(self):
+        meta = get_ai_metadata(SAMPLES_DIR / "chatgpt-1.png")
+        assert "claim_generator" in meta
+        assert "OpenAI" in meta["issuer"]
+        assert "OpenAI" in meta["synthid_watermark"]
+        assert "trainedAlgorithmicMedia" in meta["source_type"]
+
+
+@pytest.mark.parametrize(
+    "marker",
+    [
+        b"trainedAlgorithmicMedia",
+        b"compositeSynthetic",
+        b"algorithmicMedia",
+        b"compositeWithTrainedAlgorithmicMedia",
+    ],
+)
+def test_has_ai_metadata_detects_each_iptc_marker(tmp_path: Path, marker: bytes):
+    """Each IPTC digitalSourceType AI marker in XMP triggers detection."""
+    path = tmp_path / "iptc.jpg"
+    path.write_bytes(b"\xff\xd8\xff\xe1<x:xmpmeta>" + marker + b"</x:xmpmeta>\xff\xd9")
+    assert has_ai_metadata(path)
+
+
+# ── SynthID-source detection (metadata proxy) ────────────────────────
+
+
+@pytest.mark.skipif(not SAMPLES_DIR.exists(), reason="data/samples not present")
+class TestSynthIDSource:
+    """SynthID detection via the C2PA companion manifest.
+
+    Google (Imagen/Gemini) and OpenAI (ChatGPT/DALL-E/gpt-image) pair an
+    invisible SynthID pixel watermark with a C2PA manifest. Adobe Firefly and
+    Microsoft Designer sign C2PA Content Credentials but do NOT use SynthID,
+    so the discriminating signal is the C2PA *issuer*, not the mere presence
+    of a manifest. These tests run against real, committed sample images.
+    """
+
+    def test_openai_chatgpt_is_synthid_source(self):
+        assert synthid_source(SAMPLES_DIR / "chatgpt-1.png") == "OpenAI"
+
+    def test_openai_verdict_in_get_ai_metadata(self):
+        meta = get_ai_metadata(SAMPLES_DIR / "chatgpt-1.png")
+        assert "synthid_watermark" in meta
+        assert "OpenAI" in meta["synthid_watermark"]
+
+    def test_adobe_firefly_is_not_synthid_source(self):
+        # Adobe signs C2PA (trainedAlgorithmicMedia) but embeds no SynthID.
+        assert synthid_source(SAMPLES_DIR / "firefly-1.png") is None
+        assert "synthid_watermark" not in get_ai_metadata(SAMPLES_DIR / "firefly-1.png")
+
+    def test_non_ai_image_is_not_synthid_source(self):
+        assert synthid_source(SAMPLES_DIR / "not-ai-1.jpeg") is None
+
+
+class TestSynthIDSourceNonPng:
+    """SynthID-source detection must work beyond PNG.
+
+    ChatGPT/Gemini images saved as JPEG/WebP/AVIF carry their C2PA manifest in
+    a non-PNG container (JPEG APP11, ISOBMFF uuid box), so the PNG caBX parser
+    misses them. These use synthetic byte blobs (real fixtures aren't shipped).
+    """
+
+    def _c2pa_jpeg(self, tmp_path: Path, name: str, issuer: bytes, marker: bytes = b"trainedAlgorithmicMedia") -> Path:
+        path = tmp_path / name
+        # Minimal JPEG shell with an embedded C2PA-ish blob.
+        blob = b"jumbc2pa" + issuer + b"..." + marker
+        path.write_bytes(b"\xff\xd8\xff\xe1" + blob + b"\xff\xd9")
+        return path
+
+    def test_openai_c2pa_in_jpeg(self, tmp_path: Path):
+        path = self._c2pa_jpeg(tmp_path, "chatgpt.jpg", b"OpenAI")
+        assert synthid_source(path) == "OpenAI"
+
+    def test_google_c2pa_in_jpeg(self, tmp_path: Path):
+        path = self._c2pa_jpeg(tmp_path, "gemini.jpg", b"Google")
+        assert synthid_source(path) == "Google LLC"
+
+    def test_adobe_c2pa_in_jpeg_is_none(self, tmp_path: Path):
+        # Adobe signs C2PA but embeds no SynthID.
+        path = self._c2pa_jpeg(tmp_path, "firefly.jpg", b"Adobe")
+        assert synthid_source(path) is None
+
+    def test_openai_without_ai_marker_is_none(self, tmp_path: Path):
+        # Issuer present but no AI digital-source marker -> not a SynthID source.
+        path = self._c2pa_jpeg(tmp_path, "edited.jpg", b"OpenAI", marker=b"")
+        assert synthid_source(path) is None
+

 # ── remove_ai_metadata ──────────────────────────────────────────────

@@ -217,3 +334,59 @@ class TestRemoveAiMetadata:
        result = remove_ai_metadata(tmp_clean_png, output)
        assert isinstance(result, Path)
        assert result == output
+
+
+def _img_with_software(tmp_path: Path, fmt: str, software: str) -> Path:
+    """Write a tiny image carrying an EXIF Software tag."""
+    exif = piexif.dump({"0th": {piexif.ImageIFD.Software: software.encode()}, "Exif": {}, "GPS": {}, "1st": {}})
+    path = tmp_path / f"img.{fmt}"
+    Image.new("RGB", (64, 64), (100, 90, 80)).save(path, exif=exif)
+    return path
+
+
+class TestExifGenerator:
+    """exif_generator extracts AI-tool names from EXIF/XMP across formats."""
+
+    def test_avif_software_ai_tool_detected(self, tmp_path: Path):
+        path = _img_with_software(tmp_path, "avif", "Adobe Firefly")
+        assert exif_generator(path) == "Adobe Firefly"
+
+    def test_jpeg_software_ai_tool_detected(self, tmp_path: Path):
+        path = _img_with_software(tmp_path, "jpg", "ComfyUI v1.2")
+        result = exif_generator(path)
+        assert result is not None
+        assert "ComfyUI" in result
+
+    def test_plain_editor_not_flagged(self, tmp_path: Path):
+        # An ordinary editor tag carries no AI token and must not be flagged.
+        path = _img_with_software(tmp_path, "jpg", "Adobe Photoshop 25.0")
+        assert exif_generator(path) is None
+
+    def test_make_tag_ai_tool_detected(self, tmp_path: Path):
+        # Ideogram tags its output with EXIF Make="Ideogram AI" (verified on a
+        # real download), so the Make tag must be read too.
+        exif = piexif.dump({"0th": {piexif.ImageIFD.Make: b"Ideogram AI"}, "Exif": {}, "GPS": {}, "1st": {}})
+        path = tmp_path / "ideogram.jpg"
+        Image.new("RGB", (64, 64)).save(path, exif=exif)
+        assert exif_generator(path) == "Ideogram AI"
+
+    def test_camera_make_not_flagged(self, tmp_path: Path):
+        # A real camera Make ("Apple") carries no AI token -> not flagged.
+        exif = piexif.dump({"0th": {piexif.ImageIFD.Make: b"Apple"}, "Exif": {}, "GPS": {}, "1st": {}})
+        path = tmp_path / "iphone.jpg"
+        Image.new("RGB", (64, 64)).save(path, exif=exif)
+        assert exif_generator(path) is None
+
+    def test_xmp_creatortool_scan_covers_unopenable(self, tmp_path: Path):
+        # PIL can't open this fake HEIF; the raw XMP CreatorTool scan still works.
+        path = tmp_path / "fake.heic"
+        path.write_bytes(
+            b"\x00\x00\x00\x18ftypheic\x00\x00\x00\x00"
+            b"<x:xmpmeta><xmp:CreatorTool>Midjourney v7</xmp:CreatorTool></x:xmpmeta>"
+        )
+        result = exif_generator(path)
+        assert result is not None
+        assert "Midjourney" in result
+
+    def test_clean_image_is_none(self, tmp_clean_png: Path):
+        assert exif_generator(tmp_clean_png) is None
@@ -2,10 +2,19 @@

 from __future__ import annotations

+import struct
+from pathlib import Path
+
+import pytest
+
 from remove_ai_watermarks.noai.c2pa import (
+    _cbor_text_after,
+    _parse_c2pa_chunk,
    extract_c2pa_chunk,
    extract_c2pa_info,
    has_c2pa_metadata,
+    inject_c2pa_chunk,
+    synthid_verdict,
 )
 from remove_ai_watermarks.noai.cleaner import (
    has_ai_content,
@@ -26,6 +35,10 @@ from remove_ai_watermarks.noai.extractor import (
    get_ai_metadata_summary,
    has_ai_metadata,
 )
+from remove_ai_watermarks.noai.isobmff import (
+    is_isobmff,
+    strip_c2pa_boxes,
+)

 # ── Constants ───────────────────────────────────────────────────────

@@ -128,3 +141,156 @@ class TestC2PA:

    def test_c2pa_returns_false_for_non_png(self, tmp_jpeg_path):
        assert not has_c2pa_metadata(tmp_jpeg_path)
+
+
+SAMPLES_DIR = Path(__file__).resolve().parent.parent / "data" / "samples"
+
+
+@pytest.mark.skipif(not SAMPLES_DIR.exists(), reason="data/samples not present")
+class TestC2PARealSamples:
+    """Parser behavior on real committed C2PA images."""
+
+    def test_detects_c2pa_in_openai_png(self):
+        assert has_c2pa_metadata(SAMPLES_DIR / "chatgpt-1.png")
+
+    def test_extract_info_openai_fields(self):
+        info = extract_c2pa_info(SAMPLES_DIR / "chatgpt-1.png")
+        assert info["has_c2pa"] is True
+        assert "OpenAI" in info["issuer"]
+        assert "c2pa_manifest" in info  # "C2PA manifest (N bytes)"
+        assert "trainedAlgorithmicMedia" in info["source_type"]
+        # CBOR-clean claim generator, no regex artifacts (e.g. "fGPT-4o").
+        assert info["claim_generator"]
+        assert not info["claim_generator"].startswith("f")
+        assert "synthid_watermark" in info
+
+    def test_extract_info_adobe_has_no_synthid(self):
+        info = extract_c2pa_info(SAMPLES_DIR / "firefly-1.png")
+        assert "Adobe" in info["issuer"]
+        assert "synthid_watermark" not in info
+
+    def test_extract_chunk_returns_bytes(self):
+        chunk = extract_c2pa_chunk(SAMPLES_DIR / "chatgpt-1.png")
+        assert chunk is not None
+        assert chunk[4:8] == b"caBX"  # chunk type in the 8-byte header
+
+    def test_inject_round_trip(self, tmp_clean_png, tmp_path):
+        """Extract a real C2PA chunk, inject into a clean PNG, re-detect."""
+        chunk = extract_c2pa_chunk(SAMPLES_DIR / "chatgpt-1.png")
+        out = tmp_path / "injected.png"
+        inject_c2pa_chunk(tmp_clean_png, out, chunk)
+        assert has_c2pa_metadata(out)
+        assert "OpenAI" in extract_c2pa_info(out)["issuer"]
+
+
+class TestC2PAInjectValidation:
+    def test_inject_rejects_non_png(self, tmp_path):
+        with pytest.raises(ValueError, match="only supported for PNG"):
+            inject_c2pa_chunk(tmp_path / "in.jpg", tmp_path / "out.png", b"")
+
+
+# ── CBOR text extraction (parser internals) ─────────────────────────
+
+
+class TestCborTextAfter:
+    """_cbor_text_after handles the three CBOR text-string length prefixes."""
+
+    def test_direct_length(self):
+        # major-type 3, direct length (0x60 + len). "abc" -> 0x63.
+        payload = b"name" + bytes([0x63]) + b"abc"
+        assert _cbor_text_after(payload, b"name") == "abc"
+
+    def test_one_byte_length(self):
+        s = b"x" * 30
+        payload = b"name" + bytes([0x78, 30]) + s
+        assert _cbor_text_after(payload, b"name") == "x" * 30
+
+    def test_two_byte_length(self):
+        s = b"y" * 300
+        payload = b"name" + bytes([0x79]) + struct.pack(">H", 300) + s
+        assert _cbor_text_after(payload, b"name") == "y" * 300
+
+    def test_key_not_found_returns_none(self):
+        assert _cbor_text_after(b"nothing here", b"name") is None
+
+    def test_key_at_end_returns_none(self):
+        assert _cbor_text_after(b"prefixname", b"name") is None
+
+    def test_invalid_head_returns_none(self):
+        # 0x00 is not a text-string head.
+        assert _cbor_text_after(b"name" + bytes([0x00]) + b"abc", b"name") is None
+
+    def test_latin1_fallback_on_invalid_utf8(self):
+        payload = b"name" + bytes([0x61]) + b"\xff"  # len 1, invalid utf-8
+        assert _cbor_text_after(payload, b"name") is not None
+
+
+class TestSynthIDVerdict:
+    def test_format(self):
+        assert synthid_verdict("OpenAI") == "likely present (OpenAI embeds SynthID with C2PA)"
+
+    def test_multiple_vendors(self):
+        assert "Google LLC, OpenAI" in synthid_verdict("Google LLC, OpenAI")
+
+
+class TestParseChunkGuards:
+    """_parse_c2pa_chunk rejects non-printable claim_generator garbage.
+
+    On some manifests (observed: Microsoft Designer) the first ``name`` key
+    precedes a binary hash field, not the generator string. The clean issuer +
+    SynthID verdict must still come through.
+    """
+
+    def test_clean_generator_kept(self):
+        # "name" + CBOR text-string (head 0x69 = 0x60+9) "gpt-image"
+        chunk = b"...name" + bytes([0x69]) + b"gpt-image" + b"OpenAI trainedAlgorithmicMedia"
+        info: dict = {}
+        _parse_c2pa_chunk(chunk, info)
+        assert info["claim_generator"] == "gpt-image"
+        assert "OpenAI" in info["issuer"]
+        assert "synthid_watermark" in info  # OpenAI + trainedAlgorithmicMedia
+
+    def test_nonprintable_generator_dropped(self):
+        # "name" + CBOR string (head 0x64 = len 4) with a control byte -> garbage
+        chunk = b"...name" + bytes([0x64]) + b"\x81abc" + b"OpenAI trainedAlgorithmicMedia"
+        info: dict = {}
+        _parse_c2pa_chunk(chunk, info)
+        assert "claim_generator" not in info  # control-char garbage rejected
+        assert "OpenAI" in info["issuer"]  # issuer byte-search still robust
+
+
+# ── ISOBMFF (AVIF / HEIF / JPEG-XL container stripping) ──────────────
+
+FTYP = b"\x00\x00\x00\x18ftypavif\x00\x00\x00\x00avifmif1"  # 24-byte ftyp box
+
+
+class TestISOBMFF:
+    def test_is_isobmff_true(self):
+        assert is_isobmff(FTYP)
+
+    def test_is_isobmff_false_for_png(self):
+        assert not is_isobmff(b"\x89PNG\r\n\x1a\n\x00\x00")
+
+    def test_is_isobmff_false_for_short(self):
+        assert not is_isobmff(b"abc")
+
+    def test_strips_jpegxl_jumb_box(self):
+        """JPEG-XL stores JUMBF in a ``jumb`` box, always stripped."""
+        jumb = struct.pack(">I", 8 + 5) + b"jumb" + b"hello"
+        cleaned, stripped = strip_c2pa_boxes(FTYP + jumb)
+        assert stripped == 1
+        assert cleaned == FTYP
+
+    def test_keeps_non_c2pa_box_with_64bit_size(self):
+        """size==1 means a 64-bit largesize follows; non-C2PA box is kept."""
+        payload = b"\x00" * 8
+        box = b"\x00\x00\x00\x01" + b"free" + struct.pack(">Q", 16 + len(payload)) + payload
+        cleaned, stripped = strip_c2pa_boxes(FTYP + box)
+        assert stripped == 0
+        assert cleaned == FTYP + box
+
+    def test_malformed_box_does_not_crash(self):
+        # A box claiming size 4 (< 8-byte header) must terminate iteration safely.
+        cleaned, stripped = strip_c2pa_boxes(FTYP + b"\x00\x00\x00\x04XXXX")
+        assert stripped == 0
+        assert cleaned.startswith(FTYP)
@@ -87,6 +87,24 @@ class TestModelProfiles:
    def test_recommended_strength_medium(self):
        assert get_recommended_strength("unknown_type") == 0.35

+    @pytest.mark.parametrize("wm_type", ["stegastamp", "stegasamp", "treering", "ringid"])
+    def test_high_perturbation_watermark_types(self, wm_type):
+        """Robust spatial watermarks need aggressive (0.7) regeneration."""
+        assert get_recommended_strength(wm_type) == 0.7
+
+    @pytest.mark.parametrize("wm_type", ["stablesignature", "dwtectsvd", "rivagan", "ssl", "hidden"])
+    def test_low_perturbation_watermark_types(self, wm_type):
+        """Fragile frequency/latent watermarks break at low (0.04) strength."""
+        assert get_recommended_strength(wm_type) == 0.04
+
+    def test_strength_match_is_case_insensitive(self):
+        assert get_recommended_strength("TreeRing") == 0.7
+        assert get_recommended_strength("StableSignature") == 0.04
+
+    def test_strength_matches_substring_in_descriptive_name(self):
+        # e.g. a CLI passing "treering_v2" or "synthid-stegastamp" still maps.
+        assert get_recommended_strength("treering_v2") == 0.7
+

 # ── Format utilities ────────────────────────────────────────────────

@@ -0,0 +1,66 @@
+"""Tests for the SynthID corpus ingestion script."""
+
+from __future__ import annotations
+
+import csv
+import sys
+from pathlib import Path
+
+import pytest
+from click.testing import CliRunner
+
+# scripts/ is not an installed package; add it to the path for import.
+sys.path.insert(0, str(Path(__file__).resolve().parent.parent / "scripts"))
+
+import synthid_corpus
+
+SAMPLES_DIR = Path(__file__).resolve().parent.parent / "data" / "samples"
+
+
+def _manifest_rows(root: Path) -> list[dict[str, str]]:
+    with open(root / "manifest.csv", newline="") as f:
+        return list(csv.DictReader(f))
+
+
+@pytest.mark.skipif(not SAMPLES_DIR.exists(), reason="data/samples not present")
+class TestIngest:
+    def test_ingest_openai_flags_synthid_metadata(self, tmp_path: Path):
+        runner = CliRunner()
+        result = runner.invoke(
+            synthid_corpus.cli,
+            ["ingest", str(SAMPLES_DIR / "chatgpt-1.png"), "--label", "pos", "--root", str(tmp_path)],
+        )
+        assert result.exit_code == 0, result.output
+
+        rows = _manifest_rows(tmp_path)
+        assert len(rows) == 1
+        row = rows[0]
+        assert row["label"] == "pos"
+        assert row["synthid_metadata"] == "yes"
+        assert int(row["width"]) > 0
+        assert int(row["height"]) > 0
+        # The copied file lands under images/pos/ with a sha-prefixed name.
+        assert (tmp_path / "images" / "pos" / row["filename"]).exists()
+
+    def test_ingest_firefly_not_flagged(self, tmp_path: Path):
+        runner = CliRunner()
+        runner.invoke(
+            synthid_corpus.cli,
+            ["ingest", str(SAMPLES_DIR / "firefly-1.png"), "--label", "neg", "--root", str(tmp_path)],
+        )
+        rows = _manifest_rows(tmp_path)
+        assert len(rows) == 1
+        assert rows[0]["synthid_metadata"] == ""  # Adobe signs C2PA but not SynthID
+
+    def test_ingest_dedupes_by_sha256(self, tmp_path: Path):
+        runner = CliRunner()
+        args = ["ingest", str(SAMPLES_DIR / "chatgpt-1.png"), "--label", "pos", "--root", str(tmp_path)]
+        runner.invoke(synthid_corpus.cli, args)
+        runner.invoke(synthid_corpus.cli, args)  # second time: duplicate
+        assert len(_manifest_rows(tmp_path)) == 1
+
+    def test_status_on_empty_corpus(self, tmp_path: Path):
+        runner = CliRunner()
+        result = runner.invoke(synthid_corpus.cli, ["status", "--root", str(tmp_path)])
+        assert result.exit_code == 0
+        assert "empty" in result.output.lower()
@@ -0,0 +1,104 @@
+"""Tests for the SynthID pixel-carrier probe.
+
+The probe's logic is validated synthetically: a fixed carrier injected into
+uniform fills must correlate strongly, random noise must not, and simulated
+removal (dropping the carrier) must collapse the correlation. No real SynthID
+fills or GPU are needed.
+"""
+
+from __future__ import annotations
+
+import sys
+from pathlib import Path
+
+import numpy as np
+from click.testing import CliRunner
+from PIL import Image
+
+# scripts/ is not an installed package; add it to the path for import.
+sys.path.insert(0, str(Path(__file__).resolve().parent.parent / "scripts"))
+
+import synthid_pixel_probe as probe
+
+_SHAPE = (64, 64)
+
+
+def _fixed_carrier(seed: int = 7) -> np.ndarray:
+    """A reproducible 2-D carrier pattern (the stand-in for the SynthID signal)."""
+    rng = np.random.default_rng(seed)
+    return rng.standard_normal(_SHAPE)
+
+
+def _fill_with_carrier(base: float, pattern: np.ndarray, noise_seed: int, noise: float = 0.05) -> np.ndarray:
+    """A solid fill at brightness ``base`` carrying ``pattern`` plus light noise."""
+    rng = np.random.default_rng(noise_seed)
+    return base + pattern + noise * rng.standard_normal(_SHAPE)
+
+
+class TestCarrier:
+    def test_flat_image_has_zero_carrier(self):
+        # A perfectly uniform fill (std 0, like the synthetic refs) carries nothing.
+        flat = np.zeros(_SHAPE)
+        assert not np.any(probe.carrier(flat))
+
+    def test_carrier_is_unit_norm(self):
+        c = probe.carrier(_fixed_carrier() + 100.0)
+        assert np.isclose(np.linalg.norm(c), 1.0)
+
+    def test_ncc_self_is_one(self):
+        c = probe.carrier(_fixed_carrier())
+        assert np.isclose(probe.ncc(c, c), 1.0)
+
+    def test_ncc_mismatched_shape_is_zero(self):
+        a = probe.carrier(np.random.default_rng(1).standard_normal((8, 8)))
+        b = probe.carrier(np.random.default_rng(2).standard_normal((16, 16)))
+        assert probe.ncc(a, b) == 0.0
+
+
+class TestConsistency:
+    def test_shared_carrier_correlates_high(self):
+        pattern = _fixed_carrier()
+        carriers = [probe.carrier(_fill_with_carrier(b, pattern, s)) for s, b in enumerate((10, 60, 120, 200))]
+        assert probe.mean_pairwise_ncc(carriers) > 0.8
+
+    def test_random_fills_near_zero(self):
+        rng = np.random.default_rng(0)
+        carriers = [probe.carrier(rng.standard_normal(_SHAPE)) for _ in range(5)]
+        assert abs(probe.mean_pairwise_ncc(carriers)) < 0.2
+
+    def test_random_baseline_near_zero(self):
+        assert abs(probe.random_baseline(_SHAPE, 6)) < 0.2
+
+
+class TestRemoval:
+    def test_removed_carrier_collapses_correlation(self):
+        pattern = _fixed_carrier()
+        pos = [probe.carrier(_fill_with_carrier(b, pattern, s)) for s, b in enumerate((20, 90, 160))]
+        tmpl = probe.template(pos)
+        # "Cleaned" fills keep the base + noise but lose the shared pattern.
+        rng = np.random.default_rng(99)
+        cleaned = [probe.carrier(b + 0.05 * rng.standard_normal(_SHAPE)) for b in (20, 90, 160)]
+        pos_corr = float(np.mean([probe.ncc(c, tmpl) for c in pos]))
+        cleaned_corr = float(np.mean([probe.ncc(c, tmpl) for c in cleaned]))
+        assert pos_corr > 0.8
+        assert cleaned_corr < 0.2
+
+
+class TestCli:
+    def _solid_fill_png(self, tmp_path: Path, name: str, base: int, pattern: np.ndarray, seed: int) -> Path:
+        arr = np.clip(_fill_with_carrier(base, pattern, seed), 0, 255).astype(np.uint8)
+        path = tmp_path / name
+        Image.fromarray(np.stack([arr] * 3, axis=2)).save(path)
+        return path
+
+    def test_consistency_command_runs(self, tmp_path: Path):
+        pattern = _fixed_carrier()
+        paths = [str(self._solid_fill_png(tmp_path, f"f{i}.png", b, pattern, i)) for i, b in enumerate((40, 120, 200))]
+        result = CliRunner().invoke(probe.cli, ["consistency", *paths])
+        assert result.exit_code == 0, result.output
+        assert "mean pairwise NCC" in result.output
+
+    def test_consistency_needs_two_images(self, tmp_path: Path):
+        path = str(self._solid_fill_png(tmp_path, "only.png", 100, _fixed_carrier(), 0))
+        result = CliRunner().invoke(probe.cli, ["consistency", path])
+        assert result.exit_code != 0
@@ -857,6 +857,23 @@ wheels = [
    { url = "https://files.pythonhosted.org/packages/cb/b1/3846dd7f199d53cb17f49cba7e651e9ce294d8497c8c150530ed11865bb8/iniconfig-2.3.0-py3-none-any.whl", hash = "sha256:f631c04d2c48c52b84d0d0549c99ff3859c98df65b3101406327ecc7d53fbf12", size = 7484, upload-time = "2025-10-18T21:55:41.639Z" },
 ]

+[[package]]
+name = "invisible-watermark"
+version = "0.2.0"
+source = { registry = "https://pypi.org/simple" }
+dependencies = [
+    { name = "numpy", version = "2.2.6", source = { registry = "https://pypi.org/simple" }, marker = "python_full_version < '3.11'" },
+    { name = "numpy", version = "2.4.4", source = { registry = "https://pypi.org/simple" }, marker = "python_full_version >= '3.11'" },
+    { name = "opencv-python" },
+    { name = "pillow" },
+    { name = "pywavelets", version = "1.8.0", source = { registry = "https://pypi.org/simple" }, marker = "python_full_version < '3.11'" },
+    { name = "pywavelets", version = "1.9.0", source = { registry = "https://pypi.org/simple" }, marker = "python_full_version >= '3.11'" },
+    { name = "torch" },
+]
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/2b/57/18b5a914f6d7994dd349252873169e946dc824328e9a37fd15ed836deedc/invisible_watermark-0.2.0-py3-none-any.whl", hash = "sha256:644311beed9cfe4a9a5a4a46c740f47800cef184fe2e1297f3f4542e2d992f8b", size = 1633253, upload-time = "2023-07-06T13:56:28.715Z" },
+]
+
 [[package]]
 name = "jinja2"
 version = "3.1.6"
@@ -1831,6 +1848,121 @@ wheels = [
    { url = "https://files.pythonhosted.org/packages/0b/d7/1959b9648791274998a9c3526f6d0ec8fd2233e4d4acce81bbae76b44b2a/python_dotenv-1.2.2-py3-none-any.whl", hash = "sha256:1d8214789a24de455a8b8bd8ae6fe3c6b69a5e3d64aa8a8e5d68e694bbcb285a", size = 22101, upload-time = "2026-03-01T16:00:25.09Z" },
 ]

+[[package]]
+name = "pywavelets"
+version = "1.8.0"
+source = { registry = "https://pypi.org/simple" }
+resolution-markers = [
+    "python_full_version < '3.11' and sys_platform == 'win32'",
+    "python_full_version < '3.11' and sys_platform != 'win32'",
+]
+dependencies = [
+    { name = "numpy", version = "2.2.6", source = { registry = "https://pypi.org/simple" }, marker = "python_full_version < '3.11'" },
+]
+sdist = { url = "https://files.pythonhosted.org/packages/48/45/bfaaab38545a33a9f06c61211fc3bea2e23e8a8e00fedeb8e57feda722ff/pywavelets-1.8.0.tar.gz", hash = "sha256:f3800245754840adc143cbc29534a1b8fc4b8cff6e9d403326bd52b7bb5c35aa", size = 3935274, upload-time = "2024-12-04T19:54:20.593Z" }
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/65/7e/c5e398f25c70558ca195dd4144ee004666401f6167084c1e76059d7e68d8/pywavelets-1.8.0-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:f5c86fcb203c8e61d1f3d4afbfc08d626c64e4e3708207315577264c724632bf", size = 4323291, upload-time = "2024-12-04T19:53:01.836Z" },
+    { url = "https://files.pythonhosted.org/packages/d0/d7/2fc8067c3520ce25f7632b0f47b89d1b75653cab804a42700e95126f2679/pywavelets-1.8.0-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:fafb5fa126277e1690c3d6329287122fc08e4d25a262ce126e3d81b1f5709308", size = 4291864, upload-time = "2024-12-04T19:53:04.659Z" },
+    { url = "https://files.pythonhosted.org/packages/2f/17/a868aa26e45c104613d9069f9d8ec0123687cb6945062d274f20a3992436/pywavelets-1.8.0-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:dec23dfe6d5a3f4312b12456b8c546aa90a11c1138e425a885987505f0658ae0", size = 4447532, upload-time = "2024-12-04T19:53:06.383Z" },
+    { url = "https://files.pythonhosted.org/packages/53/7a/7f5889a57177e2b1182080fc2c52236d1e03a0fad5e0b3d7c5312070c0be/pywavelets-1.8.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:880a0197e9fa108939af50a95e97c1bf9b7d3e148e0fad92ea60a9ed8c8947c0", size = 4487695, upload-time = "2024-12-04T19:53:08.84Z" },
+    { url = "https://files.pythonhosted.org/packages/f9/e6/04d76d93c158919ef0d8e1d40d1d453168305031eca6733fdc844f7cbb07/pywavelets-1.8.0-cp310-cp310-musllinux_1_2_aarch64.whl", hash = "sha256:8bfa833d08b60d0bf53a7939fbbf3d98015dd34efe89cbe4e53ced880d085fc1", size = 4473752, upload-time = "2024-12-04T19:53:10.848Z" },
+    { url = "https://files.pythonhosted.org/packages/3b/a7/42ea5bbb6055abd312e45b27d931200fd6eed5414a87ec5d62020a4c651b/pywavelets-1.8.0-cp310-cp310-musllinux_1_2_x86_64.whl", hash = "sha256:e10c3fc7f4a796e94da4bca9871be2186a7bb7a3b3536a0ca9376d84263140f0", size = 4504191, upload-time = "2024-12-04T19:53:13.912Z" },
+    { url = "https://files.pythonhosted.org/packages/8c/7e/52df87a9e77adfb12c1b8be79a2053f2eb4c2507dec96ebfd2333b15ff03/pywavelets-1.8.0-cp310-cp310-win32.whl", hash = "sha256:31baf4be6940fde72cc85663154360857ac1b93c251822deaf72bb804da95031", size = 4143794, upload-time = "2024-12-04T19:53:16.296Z" },
+    { url = "https://files.pythonhosted.org/packages/01/e2/06e08230c26049740b2773952fbb12cc7186e5df655a73b1c30ba493e864/pywavelets-1.8.0-cp310-cp310-win_amd64.whl", hash = "sha256:560c39f1ff8cb37f8b8ea4b7b6eb8a14f6926c11f5cf8c09f013a58f895ed5bc", size = 4214262, upload-time = "2024-12-04T19:53:17.998Z" },
+    { url = "https://files.pythonhosted.org/packages/6c/8a/9f8e794120b55caa1c4ae8d72696111bc408251615f351a8e54a5d8c4d4e/pywavelets-1.8.0-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:e8dd5be4faed994581a8a4b3c0169be20567a9346e523f0b57f903c8f6722bce", size = 4324170, upload-time = "2024-12-04T19:53:19.66Z" },
+    { url = "https://files.pythonhosted.org/packages/3e/b8/f6246be5c78e9fa73fcbba9ab4cbfe0d4dcb79ea5491f28d673a53466134/pywavelets-1.8.0-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:8d8abaf7c120b151ef309c9ff57e0a44ba9febf49045056dbc1577526ecec6c8", size = 4294254, upload-time = "2024-12-04T19:53:21.767Z" },
+    { url = "https://files.pythonhosted.org/packages/2c/dc/ba1f212e9b43117ed28e0fd092e72e817790427400f88937ea742d260153/pywavelets-1.8.0-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:4b43a4c58707b1e8d941bec7f1d83e67c482278575ff0db3189d5c0dfae23a57", size = 4447178, upload-time = "2024-12-04T19:53:23.525Z" },
+    { url = "https://files.pythonhosted.org/packages/58/10/e59c162a11d2fedb4454abbf7b74a52390aba5edc9605bf829bfa8708dac/pywavelets-1.8.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:1c1aad0b97714e3079a2bfe48e4fb8ccd60778d0427e9ee5e0a9ff922e6c61e4", size = 4486799, upload-time = "2024-12-04T19:53:25.238Z" },
+    { url = "https://files.pythonhosted.org/packages/03/ee/90c3d0a0a3bda74e6e097e4c06bff9446ff2a4c90b8617aaf4902c46966b/pywavelets-1.8.0-cp311-cp311-musllinux_1_2_aarch64.whl", hash = "sha256:a0e1db96dcf3ce08156859df8b359e9ff66fa15061a1b90e70e020bf4cd077a0", size = 4486403, upload-time = "2024-12-04T19:53:26.954Z" },
+    { url = "https://files.pythonhosted.org/packages/05/54/58b87f8b636a9f044f3f9814d2ec696cf25f3b33af97c11811f13c364085/pywavelets-1.8.0-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:e62c8fb52ab0e8ff212fff9acae681a8f12d68b76c36fe24cc48809d5b6825ba", size = 4515011, upload-time = "2024-12-04T19:53:28.832Z" },
+    { url = "https://files.pythonhosted.org/packages/a1/d0/f755cee11ff20668114942d0e777e2b502a8e4665e1fdb2553b587aac637/pywavelets-1.8.0-cp311-cp311-win32.whl", hash = "sha256:bf327528d10de471b04bb725c4e10677fac5a49e13d41bf0d0b3a1f6d7097abf", size = 4139934, upload-time = "2024-12-04T19:53:31.421Z" },
+    { url = "https://files.pythonhosted.org/packages/7b/0b/f4b92d4f00565280ea3e62a8e3dc81a667d67ed7bd59232f2f18d55f9aff/pywavelets-1.8.0-cp311-cp311-win_amd64.whl", hash = "sha256:3814d354dd109e244ffaac3d480d29a5202212fe24570c920268237c8d276f95", size = 4214321, upload-time = "2024-12-04T19:53:33.183Z" },
+    { url = "https://files.pythonhosted.org/packages/2d/8b/4870f11559307416470158a5aa6f61e5c2a910f1645a7a836ffae580b7ad/pywavelets-1.8.0-cp312-cp312-macosx_10_13_x86_64.whl", hash = "sha256:3f431c9e2aff1a2240765eff5e804975d0fcc24c82d6f3d4271243f228e5963b", size = 4326187, upload-time = "2024-12-04T19:53:35.19Z" },
+    { url = "https://files.pythonhosted.org/packages/c4/35/66835d889fd7fbf3119c7a9bd9d9bd567fc0bb603dfba408e9226db7cb44/pywavelets-1.8.0-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:e39b0e2314e928cb850ee89b9042733a10ea044176a495a54dc84d2c98407a51", size = 4295428, upload-time = "2024-12-04T19:53:36.962Z" },
+    { url = "https://files.pythonhosted.org/packages/63/1c/42e5130226538c70d4bbbaee00eb1bc06ec3287f7ea43d5fcf85bfc761ce/pywavelets-1.8.0-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:cae701117f5c7244b7c8d48b9e92a0289637cdc02a9c205e8be83361f0c11fae", size = 4421259, upload-time = "2024-12-04T19:53:39.119Z" },
+    { url = "https://files.pythonhosted.org/packages/6f/c5/1ce93657432e22a5debc21e8b52ec6980f819ecb7fa727bb86744224d967/pywavelets-1.8.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:649936baee933e80083788e0adc4d8bc2da7cdd8b10464d3b113475be2cc5308", size = 4447650, upload-time = "2024-12-04T19:53:41.589Z" },
+    { url = "https://files.pythonhosted.org/packages/b9/d6/b54ef30daca71824f811f9d2322a978b0a58d27674b8e3af6520f67e9ec6/pywavelets-1.8.0-cp312-cp312-musllinux_1_2_aarch64.whl", hash = "sha256:8c68e9d072c536bc646e8bdce443bb1826eeb9aa21b2cb2479a43954dea692a3", size = 4448538, upload-time = "2024-12-04T19:53:44.308Z" },
+    { url = "https://files.pythonhosted.org/packages/ce/8c/1688b790e55674667ad644262f174405c2c9873cb13e773432e78b1b33e4/pywavelets-1.8.0-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:63f67fa2ee1610445de64f746fb9c1df31980ad13d896ea2331fc3755f49b3ae", size = 4485228, upload-time = "2024-12-04T19:53:46.778Z" },
+    { url = "https://files.pythonhosted.org/packages/c9/9b/69de31c3b663dadd76d1da6bf8af68d8cefff55df8e880fe96a94bb8c9ac/pywavelets-1.8.0-cp312-cp312-win32.whl", hash = "sha256:4b3c2ab669c91e3474fd63294355487b7dd23f0b51d32f811327ddf3546f4f3d", size = 4134850, upload-time = "2024-12-04T19:53:49.101Z" },
+    { url = "https://files.pythonhosted.org/packages/1c/88/9e2aa9d5fde08bfc0fb18ffb1b5307c1ed49c24930b4147e5f48571a7251/pywavelets-1.8.0-cp312-cp312-win_amd64.whl", hash = "sha256:810a23a631da596fef7196ddec49b345b1aab13525bb58547eeebe1769edbbc1", size = 4210786, upload-time = "2024-12-04T19:53:51.546Z" },
+    { url = "https://files.pythonhosted.org/packages/94/73/7ff347d77c6bda11330565050c3346c54bc210086380abeb84e402c1c9cd/pywavelets-1.8.0-cp313-cp313-macosx_10_13_x86_64.whl", hash = "sha256:441ba45c8dff8c6916dbe706958d0d7f91da675695ca0c0d75e483f6f52d0a12", size = 4321474, upload-time = "2024-12-04T19:53:53.369Z" },
+    { url = "https://files.pythonhosted.org/packages/b0/70/c58937ebbca1aba3475ca5ee63c7bcebf09f3c93891ae5942eaec7e95707/pywavelets-1.8.0-cp313-cp313-macosx_11_0_arm64.whl", hash = "sha256:24bb282bab09349d9d128ed0536fa50fff5c2147891971a69c2c36155dfeeeac", size = 4291502, upload-time = "2024-12-04T19:53:55.396Z" },
+    { url = "https://files.pythonhosted.org/packages/da/55/87b4ad6128b2e85985908e958e856e0b680cdcc03cc490e2cc995164b13a/pywavelets-1.8.0-cp313-cp313-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:426ff3799446cb4da1db04c2084e6e58edfe24225596805665fd39c14f53dece", size = 4412669, upload-time = "2024-12-04T19:53:57.393Z" },
+    { url = "https://files.pythonhosted.org/packages/bf/1a/bfca9eab23bd7b27843b0ce95c47796033a7b2c93048315f5fc5d6ac6428/pywavelets-1.8.0-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:aa0607a9c085b8285bc0d04e33d461a6c80f8c325389221ffb1a45141861138e", size = 4454604, upload-time = "2024-12-04T19:53:59.105Z" },
+    { url = "https://files.pythonhosted.org/packages/c3/23/9ce38829f57159e812c469c4f9d7b5a16c1ba922c1802985e8c504468206/pywavelets-1.8.0-cp313-cp313-musllinux_1_2_aarch64.whl", hash = "sha256:d31c36a39110e8fcc7b1a4a11cfed7d22b610c285d3e7f4fe73ec777aa49fa39", size = 4445507, upload-time = "2024-12-04T19:54:00.78Z" },
+    { url = "https://files.pythonhosted.org/packages/e5/d2/e78a976b0600a6ef7a70f4430122d6ad11b3e1cbda3c8b3565661d094678/pywavelets-1.8.0-cp313-cp313-musllinux_1_2_x86_64.whl", hash = "sha256:fa7c68ed1e5bab23b1bafe60ccbcf709b878652d03de59e961baefa5210fcd0a", size = 4479078, upload-time = "2024-12-04T19:54:02.46Z" },
+    { url = "https://files.pythonhosted.org/packages/85/4d/1c4f870010368f3aeb0bdd72929376a1988e4a122e76545bd8c56e549c96/pywavelets-1.8.0-cp313-cp313-win32.whl", hash = "sha256:2c6b359b55d713ef683e9da1529181b865a80d759881ceb9adc1c5742e4da4d8", size = 4133763, upload-time = "2024-12-04T19:54:04.016Z" },
+    { url = "https://files.pythonhosted.org/packages/c9/4f/0a709a5732e6cf9297fc87bf545cb879997cde204115f8c0cbc296c5bdd3/pywavelets-1.8.0-cp313-cp313-win_amd64.whl", hash = "sha256:4dbebcfd55ea8a85b7fc8802d411e75337170422abf6e96019d7e46c394e80e5", size = 4209548, upload-time = "2024-12-04T19:54:06.61Z" },
+    { url = "https://files.pythonhosted.org/packages/de/2a/4cac0bba67d3bc0ad697d0680539864db0a6964c7ad953d8d9d887f360b3/pywavelets-1.8.0-cp313-cp313t-macosx_11_0_arm64.whl", hash = "sha256:2e1c79784bebeafd3715c1bea6621daa2e2e6ed37b687719322e2078fb35bb70", size = 4335183, upload-time = "2024-12-04T19:54:08.349Z" },
+    { url = "https://files.pythonhosted.org/packages/58/d1/3abe4cf34a35b09ad847f0e9a85f340c1988611222926d295fa8710659e7/pywavelets-1.8.0-cp313-cp313t-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:7f489380c95013cc8fb3ef338f6d8c1a907125db453cc4dc739e2cca06fcd8b6", size = 4454723, upload-time = "2024-12-04T19:54:11.007Z" },
+    { url = "https://files.pythonhosted.org/packages/d5/62/f05dd191232ae94e0b48509bb0ee65c9d45abf5e8f3612b09fd309b41384/pywavelets-1.8.0-cp313-cp313t-musllinux_1_2_x86_64.whl", hash = "sha256:06786201a91b5e74540f4f3c115c49a29190de2eb424823abbd3a1fd75ea3e28", size = 4472192, upload-time = "2024-12-04T19:54:12.754Z" },
+    { url = "https://files.pythonhosted.org/packages/20/6a/257c95ad1e0fd395cbccd4ffec0d01cc9b51a3bb91e67d8fa10ffebc9c72/pywavelets-1.8.0-cp313-cp313t-win32.whl", hash = "sha256:f2877fb7b58c94211257dcf364b204d6ed259146fc87d5a90bf9d93c97af6226", size = 4183968, upload-time = "2024-12-04T19:54:15.099Z" },
+    { url = "https://files.pythonhosted.org/packages/6c/58/7179fd6f87153f2e339171e8cfe9bf901398a89045eefd7a3911bb9b47ad/pywavelets-1.8.0-cp313-cp313t-win_amd64.whl", hash = "sha256:ec5d723c3335ff8aa630fd4b14097077f12cc02893c91cafd60dd7b1730e780f", size = 4265431, upload-time = "2024-12-04T19:54:16.928Z" },
+]
+
+[[package]]
+name = "pywavelets"
+version = "1.9.0"
+source = { registry = "https://pypi.org/simple" }
+resolution-markers = [
+    "python_full_version >= '3.11' and sys_platform == 'win32'",
+    "python_full_version >= '3.11' and sys_platform != 'win32'",
+]
+dependencies = [
+    { name = "numpy", version = "2.4.4", source = { registry = "https://pypi.org/simple" }, marker = "python_full_version >= '3.11'" },
+]
+sdist = { url = "https://files.pythonhosted.org/packages/5a/75/50581633d199812205ea8cdd0f6d52f12a624886b74bf1486335b67f01ff/pywavelets-1.9.0.tar.gz", hash = "sha256:148d12203377772bea452a59211d98649c8ee4a05eff019a9021853a36babdc8", size = 3938340, upload-time = "2025-08-04T16:20:04.978Z" }
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/bd/8b/ca700d0c174c3a4eec1fbb603f04374d1fed84255c2a9f487cfaa749c865/pywavelets-1.9.0-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:54662cce4d56f0d6beaa6ebd34b2960f3aa4a43c83c9098a24729e9dc20a4be2", size = 4323640, upload-time = "2025-08-04T16:18:51.683Z" },
+    { url = "https://files.pythonhosted.org/packages/b5/f3/0fa57b6407ea9c4452b0bc182141256b9481b479ffbfc9d7fdb73afe193b/pywavelets-1.9.0-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:0d8ed4b4d1eab9347e8fe0c5b45008ce5a67225ce5b05766b8b1fa923a5f8b34", size = 4294938, upload-time = "2025-08-04T16:18:53.818Z" },
+    { url = "https://files.pythonhosted.org/packages/ea/95/a998313c8459a57e488ff2b18e24be9e836aedda3aa3a1673197deeaa59a/pywavelets-1.9.0-cp311-cp311-manylinux_2_27_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:862be65481fdfecfd84c6b0ca132ba571c12697a082068921bca5b5e039f1371", size = 4472829, upload-time = "2025-08-04T16:18:55.508Z" },
+    { url = "https://files.pythonhosted.org/packages/d8/8c/f316a153f7f89d2753df8a7371d15d0faab87e709fe02715dbc297c79385/pywavelets-1.9.0-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:d76b7fa8fc500b09201d689b4f15bf5887e30ffbe2e1f338eb8470590eb4521a", size = 4524936, upload-time = "2025-08-04T16:18:57.146Z" },
+    { url = "https://files.pythonhosted.org/packages/24/f7/89fdc1caef4b384a341a8e149253e23f36c1702bbb986a26123348624854/pywavelets-1.9.0-cp311-cp311-musllinux_1_2_aarch64.whl", hash = "sha256:aa859d0b686a697c87a47e29319aebe44125f114a4f8c7e444832b921f52de5a", size = 4481475, upload-time = "2025-08-04T16:18:58.725Z" },
+    { url = "https://files.pythonhosted.org/packages/82/53/b733fbfb71853e4a5c430da56e325a763562d65241dd785f0fadb67aed6a/pywavelets-1.9.0-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:20e97b84a263003e2c7348bcf72beba96edda1a6169f072dc4e4d4ee3a6c7368", size = 4527994, upload-time = "2025-08-04T16:18:59.917Z" },
+    { url = "https://files.pythonhosted.org/packages/ed/15/5f6a6e9fdad8341e42642ed622a5f3033da4ea9d426cc3e574ae418b4726/pywavelets-1.9.0-cp311-cp311-win32.whl", hash = "sha256:f8330cdbfa506000e63e79525716df888998a76414c5cd6ecd9a7e371191fb05", size = 4136109, upload-time = "2025-08-04T16:19:01.511Z" },
+    { url = "https://files.pythonhosted.org/packages/fd/33/62dbb4aea86ec9d79b283127c42cc896f4d4ff265a9aeb1337a7836dd550/pywavelets-1.9.0-cp311-cp311-win_amd64.whl", hash = "sha256:ed10959a17df294ef55948dcc76367d59ec7b6aad67e38dd4e313d2fe3ad47b2", size = 4228321, upload-time = "2025-08-04T16:19:03.164Z" },
+    { url = "https://files.pythonhosted.org/packages/5c/37/3fda13fb2518fdd306528382d6b18c116ceafefff0a7dccd28f1034f4dd2/pywavelets-1.9.0-cp312-cp312-macosx_10_13_x86_64.whl", hash = "sha256:30baa0788317d3c938560c83fe4fc43817342d06e6c9662a440f73ba3fb25c9b", size = 4320835, upload-time = "2025-08-04T16:19:04.855Z" },
+    { url = "https://files.pythonhosted.org/packages/36/65/a5549325daafc3eae4b52de076798839eaf529a07218f8fb18cccefe76a1/pywavelets-1.9.0-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:df7436a728339696a7aa955c020ae65c85b0d9d2b5ff5b4cf4551f5d4c50f2c7", size = 4290469, upload-time = "2025-08-04T16:19:06.178Z" },
+    { url = "https://files.pythonhosted.org/packages/05/85/901bb756d37dfa56baa26ef4a3577aecfe9c55f50f51366fede322f8c91d/pywavelets-1.9.0-cp312-cp312-manylinux_2_27_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:07b26526db2476974581274c43a9c2447c917418c6bd03c8d305ad2a5cd9fac3", size = 4437717, upload-time = "2025-08-04T16:19:07.514Z" },
+    { url = "https://files.pythonhosted.org/packages/0f/34/0f54dd9c288941294898877008bcb5c07012340cc9c5db9cff1bd185d449/pywavelets-1.9.0-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:573b650805d2f3c981a0e5ae95191c781a722022c37a0f6eba3fa7eae8e0ee17", size = 4483843, upload-time = "2025-08-04T16:19:08.857Z" },
+    { url = "https://files.pythonhosted.org/packages/48/1f/cff6bb4ea64ff508d8cac3fe113c0aa95310a7446d9efa6829027cc2afdf/pywavelets-1.9.0-cp312-cp312-musllinux_1_2_aarch64.whl", hash = "sha256:3747ec804492436de6e99a7b6130480e53406d047e87dc7095ab40078a515a23", size = 4442236, upload-time = "2025-08-04T16:19:11.061Z" },
+    { url = "https://files.pythonhosted.org/packages/ce/53/a3846eeefe0fb7ca63ae045f038457aa274989a15af793c1b824138caf98/pywavelets-1.9.0-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:5163665686219c3f43fd5bbfef2391e87146813961dad0f86c62d4aed561f547", size = 4488077, upload-time = "2025-08-04T16:19:12.333Z" },
+    { url = "https://files.pythonhosted.org/packages/f7/98/44852d2fe94455b72dece2db23562145179d63186a1c971125279a1c381f/pywavelets-1.9.0-cp312-cp312-win32.whl", hash = "sha256:80b8ab99f5326a3e724f71f23ba8b0a5b03e333fa79f66e965ea7bed21d42a2f", size = 4134094, upload-time = "2025-08-04T16:19:13.564Z" },
+    { url = "https://files.pythonhosted.org/packages/2c/a7/0d9ee3fe454d606e0f5c8e3aebf99d2ecddbfb681826a29397729538c8f1/pywavelets-1.9.0-cp312-cp312-win_amd64.whl", hash = "sha256:92bfb8a117b8c8d3b72f2757a85395346fcbf37f50598880879ae72bd8e1c4b9", size = 4213900, upload-time = "2025-08-04T16:19:14.939Z" },
+    { url = "https://files.pythonhosted.org/packages/db/a7/dec4e450675d62946ad975f5b4d924437df42d2fae46e91dfddda2de0f5a/pywavelets-1.9.0-cp313-cp313-macosx_10_13_x86_64.whl", hash = "sha256:74f8455c143818e4b026fc67b27fd82f38e522701b94b8a6d1aaf3a45fcc1a25", size = 4316201, upload-time = "2025-08-04T16:19:16.259Z" },
+    { url = "https://files.pythonhosted.org/packages/aa/0c/b54b86596c0df68027e48c09210e907e628435003e77048384a2dd6767e3/pywavelets-1.9.0-cp313-cp313-macosx_11_0_arm64.whl", hash = "sha256:c50320fe0a4a23ddd8835b3dc9b53b09ee05c7cc6c56b81d0916f04fc1649070", size = 4286838, upload-time = "2025-08-04T16:19:17.92Z" },
+    { url = "https://files.pythonhosted.org/packages/5a/9c/333969c3baad8af2e7999e83addcb7bb1d1fd48e2d812fb27e2e89582cb1/pywavelets-1.9.0-cp313-cp313-manylinux_2_27_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:d6e059265223ed659e5214ab52a84883c88ddf3decbf08d7ec6abb8e4c5ed7be", size = 4430753, upload-time = "2025-08-04T16:19:19.529Z" },
+    { url = "https://files.pythonhosted.org/packages/e5/1b/a24c6ff03b026b826ad7b9267bd63cd34ce026795a0302f8a5403840b8e7/pywavelets-1.9.0-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:ae10ed46c139c7ddb8b1249cfe0989f8ccb610d93f2899507b1b1573a0e424b5", size = 4491315, upload-time = "2025-08-04T16:19:20.717Z" },
+    { url = "https://files.pythonhosted.org/packages/d7/c7/e3fbb502fca3469e51ced4f1e1326364c338be91edc5db5a8ddd26b303fa/pywavelets-1.9.0-cp313-cp313-musllinux_1_2_aarch64.whl", hash = "sha256:c8f8b1cc2df012401cb837ee6fa2f59607c7b4fe0ff409d9a4f6906daf40dc86", size = 4437654, upload-time = "2025-08-04T16:19:22.359Z" },
+    { url = "https://files.pythonhosted.org/packages/92/44/c9b25084048d9324881a19b88e0969a4141bcfdc1d218f1b4b680b7af1c1/pywavelets-1.9.0-cp313-cp313-musllinux_1_2_x86_64.whl", hash = "sha256:db43969c7a8fbb17693ecfd14f21616edc3b29f0e47a49b32fa4127c01312a67", size = 4496435, upload-time = "2025-08-04T16:19:23.842Z" },
+    { url = "https://files.pythonhosted.org/packages/cd/b6/b27ec18c72b1dee3314e297af39c5f8136d43cc130dd93cb6c178ca820e5/pywavelets-1.9.0-cp313-cp313-win32.whl", hash = "sha256:9e7d60819d87dcd6c68a2d1bc1d37deb1f4d96607799ab6a25633ea484dcda41", size = 4132709, upload-time = "2025-08-04T16:19:25.415Z" },
+    { url = "https://files.pythonhosted.org/packages/0a/87/78ef3f9fb36cdb16ee82371d22c3a7c89eeb79ec8c9daef6222060da6c79/pywavelets-1.9.0-cp313-cp313-win_amd64.whl", hash = "sha256:0d70da9d7858c869e24dc254f16a61dc09d8a224cad85a10c393b2eccddeb126", size = 4213377, upload-time = "2025-08-04T16:19:26.875Z" },
+    { url = "https://files.pythonhosted.org/packages/8b/cd/ca0d9db0ff29e3843f6af60c2f5eb588794e05ca8eeb872a595867b1f3f5/pywavelets-1.9.0-cp313-cp313t-macosx_10_13_x86_64.whl", hash = "sha256:4dc85f44c38d76a184a1aa2cb038f802c3740428c9bb877525f4be83a223b134", size = 4354336, upload-time = "2025-08-04T16:19:28.745Z" },
+    { url = "https://files.pythonhosted.org/packages/82/d6/70afefcc1139f37d02018a3b1dba3b8fc87601bb7707d9616b7f7a76e269/pywavelets-1.9.0-cp313-cp313t-macosx_11_0_arm64.whl", hash = "sha256:7acf6f950c6deaecd210fbff44421f234a8ca81eb6f4da945228e498361afa9d", size = 4335721, upload-time = "2025-08-04T16:19:30.371Z" },
+    { url = "https://files.pythonhosted.org/packages/cd/3a/713f731b9ed6df0c36269c8fb62be8bb28eb343b9e26b13d6abda37bce38/pywavelets-1.9.0-cp313-cp313t-manylinux_2_27_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:144d4fc15c98da56654d0dca2d391b812b8d04127b194a37ad4a497f8e887141", size = 4418702, upload-time = "2025-08-04T16:19:31.743Z" },
+    { url = "https://files.pythonhosted.org/packages/44/e8/f801eb4b5f7a316ba20054948c5d6b27b879c77fab2674942e779974bd86/pywavelets-1.9.0-cp313-cp313t-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:1aa3729585408a979d655736f74b995b511c86b9be1544f95d4a3142f8f4b8b5", size = 4470023, upload-time = "2025-08-04T16:19:32.963Z" },
+    { url = "https://files.pythonhosted.org/packages/e9/cc/44b002cb16f2a392f2082308dd470b3f033fa4925d3efa7c46f790ce895a/pywavelets-1.9.0-cp313-cp313t-musllinux_1_2_aarch64.whl", hash = "sha256:e0e24ad6b8eb399c49606dd1fcdcbf9749ad7f6d638be3fe6f59c1f3098821e2", size = 4426498, upload-time = "2025-08-04T16:19:34.151Z" },
+    { url = "https://files.pythonhosted.org/packages/91/fe/2b70276ede7878c5fe8356ca07574db5da63e222ce39a463e84bfad135e8/pywavelets-1.9.0-cp313-cp313t-musllinux_1_2_x86_64.whl", hash = "sha256:3830e6657236b53a3aae20c735cccead942bb97c54bbca9e7d07bae01645fe9c", size = 4477528, upload-time = "2025-08-04T16:19:35.932Z" },
+    { url = "https://files.pythonhosted.org/packages/e7/ed/d58b540c15e36508cfeded7b0d39493e811b0dce18d9d4e6787fb2e89685/pywavelets-1.9.0-cp313-cp313t-win32.whl", hash = "sha256:81bb65facfbd7b50dec50450516e72cdc51376ecfdd46f2e945bb89d39bfb783", size = 4186493, upload-time = "2025-08-04T16:19:37.198Z" },
+    { url = "https://files.pythonhosted.org/packages/84/b2/12a849650d618a86bbe4d8876c7e20a7afe59a8cad6f49c57eca9af26dfa/pywavelets-1.9.0-cp313-cp313t-win_amd64.whl", hash = "sha256:47d52cf35e2afded8cfe1133663f6f67106a3220b77645476ae660ad34922cb4", size = 4274821, upload-time = "2025-08-04T16:19:38.436Z" },
+    { url = "https://files.pythonhosted.org/packages/ba/1f/18c82122547c9eec2232d800b02ada1fbd30ce2136137b5738acca9d653e/pywavelets-1.9.0-cp314-cp314-macosx_10_13_x86_64.whl", hash = "sha256:53043d2f3f4e55a576f51ac594fe33181e1d096d958e01524db5070eb3825306", size = 4314440, upload-time = "2025-08-04T16:19:39.701Z" },
+    { url = "https://files.pythonhosted.org/packages/eb/e1/1c92ac6b538ef5388caf1a74af61cf6af16ea6d14115bb53357469cb38d6/pywavelets-1.9.0-cp314-cp314-macosx_11_0_arm64.whl", hash = "sha256:56bc36b42b1b125fd9cb56e7956b22f8d0f83c1093f49c77fc042135e588c799", size = 4290162, upload-time = "2025-08-04T16:19:41.322Z" },
+    { url = "https://files.pythonhosted.org/packages/96/d3/d856a2cac8069c20144598fa30a43ca40b5df2e633230848a9a942faf04a/pywavelets-1.9.0-cp314-cp314-manylinux_2_27_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:08076eb9a182ddc6054ac86868fb71df6267c341635036dc63d20bdbacd9ad7e", size = 4437162, upload-time = "2025-08-04T16:19:42.556Z" },
+    { url = "https://files.pythonhosted.org/packages/c9/54/777e0495acd4fb008791e84889be33d6e7fc8af095b441d939390b7d2491/pywavelets-1.9.0-cp314-cp314-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:4ee1ee7d80f88c64b8ec3b5021dd1e94545cc97f0cd479fb51aa7b10f6def08e", size = 4498169, upload-time = "2025-08-04T16:19:43.791Z" },
+    { url = "https://files.pythonhosted.org/packages/76/68/81b97f4d18491a18fbe17e06e2eee80a591ce445942f7b6f522de07813c5/pywavelets-1.9.0-cp314-cp314-musllinux_1_2_aarch64.whl", hash = "sha256:3226b6f62838a6ccd7782cb7449ee5d8b9d61999506c1d9b03b2baf41b01b6fd", size = 4443318, upload-time = "2025-08-04T16:19:45.368Z" },
+    { url = "https://files.pythonhosted.org/packages/92/74/5147f2f0436f7aa131cb1bc13dba32ef5f3862748ae1c7366b4cde380362/pywavelets-1.9.0-cp314-cp314-musllinux_1_2_x86_64.whl", hash = "sha256:9fb7f4b11d18e2db6dd8deee7b3ce8343d45f195f3f278c2af6e3724b1b93a24", size = 4503294, upload-time = "2025-08-04T16:19:46.632Z" },
+    { url = "https://files.pythonhosted.org/packages/3d/d4/af998cc71e869919e0ab45471bd43e91d055ac7bc3ce6f56cc792c9b6bc8/pywavelets-1.9.0-cp314-cp314-win32.whl", hash = "sha256:9902d9fc9812588ab2dce359a1307d8e7f002b53a835640e2c9388fe62a82fd4", size = 4144478, upload-time = "2025-08-04T16:19:47.974Z" },
+    { url = "https://files.pythonhosted.org/packages/7d/66/1d071eae5cc3e3ad0e45334462f8ce526a79767ccb759eb851aa5b78a73a/pywavelets-1.9.0-cp314-cp314-win_amd64.whl", hash = "sha256:7e57792bde40e331d6cc65458e5970fd814dba18cfc4e9add9d051e901a7b7c7", size = 4227186, upload-time = "2025-08-04T16:19:49.57Z" },
+    { url = "https://files.pythonhosted.org/packages/bf/1f/da0c03ac99bd9d20409c0acf6417806d4cf333d70621da9f535dd0cf27fa/pywavelets-1.9.0-cp314-cp314t-macosx_10_13_x86_64.whl", hash = "sha256:b47c72fb4b76d665c4c598a5b621b505944e5b761bf03df9d169029aafcb652f", size = 4354391, upload-time = "2025-08-04T16:19:51.221Z" },
+    { url = "https://files.pythonhosted.org/packages/95/b6/de9e225d8cc307fbb4fda88aefa79442775d5e27c58ee4d3c8a8580ceba6/pywavelets-1.9.0-cp314-cp314t-macosx_11_0_arm64.whl", hash = "sha256:969e369899e7eab546ea5d77074e4125082e6f9dad71966499bf5dee3758be55", size = 4335810, upload-time = "2025-08-04T16:19:52.813Z" },
+    { url = "https://files.pythonhosted.org/packages/33/3b/336761359d07cd44a4233ca854704ff2a9e78d285879ccc82d254b9daa57/pywavelets-1.9.0-cp314-cp314t-manylinux_2_27_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:8aeffd4f35036c1fade972a61454de5709a7a8fc9a7d177eefe3ac34d76962e5", size = 4422220, upload-time = "2025-08-04T16:19:54.068Z" },
+    { url = "https://files.pythonhosted.org/packages/98/61/76ccc7ada127f14f65eda40e37407b344fd3713acfca7a94d7f0f67fe57d/pywavelets-1.9.0-cp314-cp314t-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:f63f400fcd4e7007529bd06a5886009760da35cd7e76bb6adb5a5fbee4ffeb8c", size = 4470156, upload-time = "2025-08-04T16:19:55.379Z" },
+    { url = "https://files.pythonhosted.org/packages/e0/de/142ca27ee729cf64113c2560748fcf2bd45b899ff282d6f6f3c0e7f177bb/pywavelets-1.9.0-cp314-cp314t-musllinux_1_2_aarch64.whl", hash = "sha256:a63bcb6b5759a7eb187aeb5e8cd316b7adab7de1f4b5a0446c9a6bcebdfc22fb", size = 4430167, upload-time = "2025-08-04T16:19:56.566Z" },
+    { url = "https://files.pythonhosted.org/packages/ca/5e/90b39adff710d698c00ba9c3125e2bec99dad7c5f1a3ba37c73a78a6689f/pywavelets-1.9.0-cp314-cp314t-musllinux_1_2_x86_64.whl", hash = "sha256:9950eb7c8b942e9bfa53d87c7e45a420dcddbd835c4c5f1aca045a3f775c6113", size = 4477378, upload-time = "2025-08-04T16:19:58.162Z" },
+    { url = "https://files.pythonhosted.org/packages/f1/1a/89f5f4ebcb9d34d9b7b2ac0a868c8b6d8c78d699a36f54407a060cea0566/pywavelets-1.9.0-cp314-cp314t-win32.whl", hash = "sha256:097f157e07858a1eb370e0d9c1bd11185acdece5cca10756d6c3c7b35b52771a", size = 4209132, upload-time = "2025-08-04T16:20:00.371Z" },
+    { url = "https://files.pythonhosted.org/packages/68/d2/a8065103f5e2e613b916489e6c85af6402a1ec64f346d1429e2d32cb8d03/pywavelets-1.9.0-cp314-cp314t-win_amd64.whl", hash = "sha256:3b6ff6ba4f625d8c955f68c2c39b0a913776d406ab31ee4057f34ad4019fb33b", size = 4306793, upload-time = "2025-08-04T16:20:02.934Z" },
+]
+
 [[package]]
 name = "pyyaml"
 version = "6.0.3"
@@ -2018,7 +2150,7 @@ wheels = [

 [[package]]
 name = "remove-ai-watermarks"
-version = "0.4.1"
+version = "0.5.0"
 source = { editable = "." }
 dependencies = [
    { name = "click" },
@@ -2037,6 +2169,7 @@ all = [
    { name = "color-matcher" },
    { name = "controlnet-aux" },
    { name = "diffusers" },
+    { name = "invisible-watermark" },
    { name = "pyright" },
    { name = "pytest" },
    { name = "pytest-cov" },
@@ -2046,7 +2179,11 @@ all = [
    { name = "transformers" },
    { name = "ultralytics" },
 ]
+detect = [
+    { name = "invisible-watermark" },
+]
 dev = [
+    { name = "invisible-watermark" },
    { name = "pyright" },
    { name = "pytest" },
    { name = "pytest-cov" },
@@ -2070,6 +2207,8 @@ requires-dist = [
    { name = "color-matcher", marker = "extra == 'gpu'", specifier = ">=0.5.0" },
    { name = "controlnet-aux", marker = "extra == 'gpu'", specifier = ">=0.0.9" },
    { name = "diffusers", marker = "extra == 'gpu'", specifier = ">=0.38.0" },
+    { name = "invisible-watermark", marker = "extra == 'detect'", specifier = ">=0.2.0" },
+    { name = "invisible-watermark", marker = "extra == 'dev'", specifier = ">=0.2.0" },
    { name = "numpy", specifier = ">=1.24.0" },
    { name = "opencv-python-headless", specifier = ">=4.8.0" },
    { name = "piexif", specifier = ">=1.1.3" },
@@ -2078,7 +2217,7 @@ requires-dist = [
    { name = "pytest", marker = "extra == 'dev'", specifier = ">=8.0.0" },
    { name = "pytest-cov", marker = "extra == 'dev'", specifier = ">=4.1.0" },
    { name = "python-dotenv", specifier = ">=1.0.0" },
-    { name = "remove-ai-watermarks", extras = ["gpu", "dev"], marker = "extra == 'all'" },
+    { name = "remove-ai-watermarks", extras = ["gpu", "detect", "dev"], marker = "extra == 'all'" },
    { name = "rich", specifier = ">=13.0.0" },
    { name = "ruff", marker = "extra == 'dev'", specifier = ">=0.4.0" },
    { name = "safetensors", marker = "extra == 'gpu'" },
@@ -2086,7 +2225,7 @@ requires-dist = [
    { name = "transformers", marker = "extra == 'gpu'", specifier = ">=4.35.0" },
    { name = "ultralytics", marker = "extra == 'gpu'", specifier = ">=8.0.0" },
 ]
-provides-extras = ["gpu", "dev", "all"]
+provides-extras = ["gpu", "detect", "dev", "all"]

 [[package]]
 name = "requests"
Author	SHA1	Message	Date
test-userandClaude Opus 4.7	0762807e42	chore(release): v0.5.0 — image provenance (identify) + SD/SDXL/FLUX + EXIF/XMP detection New since v0.4.1: - identify command: aggregate C2PA, IPTC, SD/ComfyUI params, SynthID proxy, visible sparkle, open invisible watermark into one provenance verdict (--json, --no-visible). - Open SD/SDXL/FLUX invisible-watermark detection (imwatermark, extra: detect). - EXIF Software / XMP CreatorTool generator-tag reading (incl. AVIF/HEIF). - Stability AI + Microsoft/Bing C2PA issuers; SynthID metadata detection. - SynthID reference corpus + experimental pixel-carrier probe. - Fix: __version__ was stuck at 0.3.4 (banner mismatch), now synced. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-25 08:58:45 -07:00
test-userandClaude Opus 4.7	626f43aec9	docs: correct SynthID spectral-carrier understanding Deeper re-examination (2026-05-25) of github.com/aloshdenny/reverse-SynthID on our own data corrects the earlier over-stated dead-end: - The carrier IS real on solid fills -- measured via per-bin PHASE COHERENCE (the prior probe used spatial/FFT-magnitude NCC, which can't see a fixed-phase carrier). White gemini-2.5-flash fills: coherence 0.86 at carriers (0,+/-7..12,20..23) vs 0.31 random; single-image phase-match +0.83 vs -0.24 for real photos. - But it does not generalize: carriers are model-version/resolution/color specific (v4 codebook for 3.1-flash/nb-pro scores ~0.5 on 2.5-flash), and collapse on real content (coherence ~random; v4 content 0.518 vs neg 0.504, no separation). Net: a controlled-fill characterizer, not a real-content detector. Metadata proxy + visible sparkle + online oracles remain the ceiling. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-25 08:57:35 -07:00
test-userandClaude Opus 4.7	ede35a3db5	feat(metadata): read EXIF Make tag; collect Ideogram/Recraft/Krea-FLUX Collected live samples from three popular generators we lacked: - Ideogram tags its downloads with EXIF Make="Ideogram AI" (no C2PA, no SynthID, no imwatermark) -- the Make tag is its only signal. exif_generator only read Software/Artist/ImageDescription, so it missed this; now reads Make too. Real cameras put "Apple"/"Canon" in Make (no AI token), so this stays low-false-positive. 4 originals ingested. - Recraft (PNG export) and Krea hosting FLUX 2: downloads carry NO detectable signal -- no C2PA/EXIF/IPTC, and notably no imwatermark despite Krea running FLUX. identify correctly reports 'unknown'. Both ingested as neg fixtures. Lesson recorded in CLAUDE.md: the imwatermark detector fires only on pristine output from a pipeline that runs the encoder (diffusers default, official BFL), not from re-hosts (Krea/Stability) or re-encoded exports (Recraft/Canva). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-24 18:38:56 -07:00
test-userandClaude Opus 4.7	ad3b8ee248	feat(identify): read EXIF Software / XMP CreatorTool generator tags Closes the documented gap where EXIF/XMP fields inside AVIF/HEIF/JXL went unparsed. metadata.exif_generator extracts the EXIF Software/Artist tag (via PIL+piexif, which opens AVIF natively) and the XMP CreatorTool (via a container-agnostic raw-byte scan that also covers HEIF/JXL that PIL can't open), and matches against AI_GENERATOR_TOKENS so only generator names (Firefly, DALL-E, Midjourney, ComfyUI, ...) fire -- a plain 'Adobe Photoshop' or 'GIMP' tag is not flagged. identify() surfaces it as a high-confidence signal and uses it for platform attribution when no C2PA names a platform, so an AVIF/HEIF whose only AI signal is an EXIF/XMP generator tag is now caught. Validated with synthesized fixtures (the 'no positive fixtures' blocker was self-imposed): real AVIF and JPEG written with EXIF Software via PIL, plus an XMP CreatorTool raw-scan fixture. Zero false positives across the 109-image corpus (real iPhone photos carry no AI generator token). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-24 17:56:39 -07:00
test-userandClaude Opus 4.7	3a1c5427c8	feat(c2pa): recognize Stability AI issuer; fix Microsoft platform label Collected live C2PA positives from Bing Image Creator and Stability Brand Studio (DreamStudio successor) and learned two things our scan got wrong: - Bing now runs Microsoft's own MAI-Image model, not DALL-E, and signs C2PA as 'Microsoft'. The scan caught it, but the platform label claimed 'Microsoft Designer (DALL-E / OpenAI backend)'. Relabeled model-neutral: 'Microsoft (Bing Image Creator / Designer)'. - Stability signs C2PA as 'Stability AI' (cert 'Stability AI Ltd'), which was not in C2PA_ISSUERS, so it read as 'unknown signer'. Added the issuer and a platform mapping. Stability uses no SynthID and (on its current Stable Image model) no imwatermark watermark -- verified, both negative. Both ingested as SynthID-negative corpus fixtures (they are AI but not SynthID) for issuer-coverage. Canva skipped: its downloads are re-encoded design exports that strip C2PA, so a Canva sample would be inconclusive. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-24 17:12:42 -07:00
test-userandClaude Opus 4.7	27ad5b7645	feat(identify): detect open SD/SDXL/FLUX invisible watermark Research found one locally-fillable detection gap: Stable Diffusion, SDXL, and FLUX all embed an open DWT-DCT watermark via the invisible-watermark (imwatermark) library -- a PUBLIC decoder, no secret key, unlike SynthID. New invisible_watermark.py decodes the known fixed patterns (verified against upstream source: diffusers SDXL WATERMARK_MESSAGE, FLUX.2 src/flux2/watermark.py, and the 'StableDiffusionV1' default string) and identify() reports the scheme as a high-confidence signal. Verified locally end-to-end: embedding SDXL's exact 48-bit message and decoding it back recovers 48/48 bits; a clean image and our own fal-SDXL outputs decode to ~21/48 (no match). Caveat baked into the report: the watermark is fragile -- gone after JPEG q90 -- so it confirms origin only on pristine files; absence is never proof. imwatermark is an optional dep (extra 'detect'; pulls non-headless opencv), so the import is guarded and the signal is skipped when absent. CLI --no-visible now means metadata-only (skips both pixel-domain detectors). Also records the broader watermarking landscape in CLAUDE.md: which services are locally detectable (SD/SDXL/FLUX), C2PA-covered (Bing/Canva/ Getty/Shutterstock unsampled), or proprietary-only like SynthID (Amazon Titan/Nova, Kakao). Midjourney embeds neither C2PA nor an invisible mark. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-24 16:53:59 -07:00
test-userandClaude Opus 4.7	7dcc922617	feat(probe): solid-fill SynthID carrier probe; corpus reconfirms no pixel detector scripts/synthid_pixel_probe.py is an experimental/diagnostic tool for the one pixel-domain question that isn't a dead-end: on solid-color fills the zero-mean residual IS essentially the watermark carrier. Two modes: 'consistency' (mean pairwise NCC of carriers across fills vs random baseline) and 'removal' (does the pipeline drop the carrier toward baseline?). Logic validated synthetically (injected carrier correlates, random noise doesn't, simulated removal collapses it) -- no real fills or GPU needed. Running its metric on the corpus independently re-confirms the documented dead-end for real content: at matched resolution SynthID positives do not cluster apart from negatives (within-Gemini 0.07; at 1024 px pos-vs-neg >= pos-vs-pos). An apparent 0.62 among 1254px ChatGPT positives turned out to be near-duplicate content (5 renders of one prompt at ~0.92; a distinct ChatGPT image scored ~0 against them), not a shared carrier. The probe is solid-fills-only; do not use on real content. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-24 16:35:39 -07:00
test-userandClaude Opus 4.7	144b98cf0b	docs: record external AI-detector models as out of scope Generic HuggingFace AI-vs-real classifiers are per-generator, degrade off-distribution, are untested on the metadata-stripped surfaces we care about (gpt-image, Gemini Nano Banana), and our own SDXL pass would likely defeat them as it does SynthID. Detection stays local + signal-based. Decision 2026-05-24. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-24 16:27:00 -07:00
test-userandClaude Opus 4.7	1a9f3e4fe5	test(identify): cover provenance branches, CLI, sparkle helper Adds 20 tests around the new provenance path: - identify(): local SD/ComfyUI params -> local-pipeline attribution; visible-sparkle gating at the 0.5 threshold (mocked detector: above, below, unavailable, opt-out); metadata verdict not downgraded by a sparkle hit; OpenAI/SynthID caveats + dedup; ProvenanceReport is JSON-serializable (the CLI --json path); and the honest edge where a C2PA manifest without an AI source marker stays 'unknown'. - CLI 'identify': help, clean PNG, AI PNG platform, valid --json, missing file. - gemini_engine.detect_sparkle_confidence: float in range for a real image, None for an unreadable file. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-24 16:27:00 -07:00
test-userandClaude Opus 4.7	fa104bcade	feat(identify): provenance command (platform + watermark inventory) New 'identify' command and identify.py module: upload an image, get one ProvenanceReport answering where it was made and what watermarks it carries. Aggregates every locally-readable signal: - C2PA Content Credentials -> generating platform (issuer + generator). - IPTC digitalSourceType 'Made with AI' (Meta and others). - Embedded SD/ComfyUI generation parameters (local pipelines). - SynthID metadata proxy (Google / OpenAI C2PA companion). - Visible Gemini sparkle (cv2 fallback for the stripped-metadata case), promoted only at confidence >= 0.5 (corpus-tuned: Gemini sparkles score >= 0.56, non-sparkle <= 0.49). is_ai_generated is True or None, never asserted False -- stripped metadata leaves no local proof of a clean origin, so absence of signals is reported as 'unknown' with an explicit caveat. The SynthID pixel watermark remains locally undecodable; the report says so. Non-PNG containers (JPEG/WebP/AVIF/HEIF/JXL) get the same issuer + generator attribution via a binary scan (the caBX parser is PNG-only). The cv2 dependency is isolated in gemini_engine.detect_sparkle_confidence so identify.py stays type-clean. CLI supports --json and --no-visible. Validated against the 109-image corpus: 14/14 positives flagged AI, 93/94 negatives clean (the one 'neg' flagged is a Meta image that genuinely carries the IPTC tag -- correct), zero true errors. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-24 16:19:26 -07:00
test-userandClaude Opus 4.7	f36320ff39	fix(metadata): guard get_ai_metadata PIL open against non-OSError get_ai_metadata opened the file with PIL unguarded, so a HEIC (or any format PIL can't open without optional plugins) raised UnidentifiedImageError instead of falling through to the binary scan -- unlike has_ai_metadata, which already guards. Wrap the open in except Exception and continue to the C2PA/IPTC path. Regression test feeds an unopenable .heic shell and asserts no raise. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-24 16:19:15 -07:00
test-userandClaude Opus 4.7	af787fd8d6	docs(corpus): per-platform watermark map + surface-dependent blind spot Grow the SynthID corpus to 109 originals (91 iPhone-photo negatives, 2 positives) and document what was learned studying 8 platforms: - README: per-platform watermark map (C2PA issuer / SynthID pixel / IPTC / visible sparkle per platform) and an "originals, not previews" note (re-encoded previews strip metadata, so a clean preview is not proof). - CLAUDE.md: surface-dependent blind spot -- the same Google model wraps C2PA in the Gemini app but emits the SynthID pixel watermark + sparkle with no C2PA/IPTC via the API/playground (AI Studio, Nano Banana), so synthid_source returns None despite SynthID being present; only the pixel oracle or the visible-sparkle detector catches those. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-24 15:55:17 -07:00
test-userandClaude Opus 4.7	be853011f3	feat(corpus): read HEIC dimensions via macOS sips PIL cannot open iPhone HEIC without pillow-heif, so width/height stayed 0 for those negatives. Fall back to sips -g pixelWidth/pixelHeight on macOS when PIL fails; returns (0,0) elsewhere. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-24 15:55:11 -07:00
test-userandClaude Opus 4.7	6cef1d59f0	fix(c2pa): drop non-printable claim_generator garbage On some manifests (observed: Microsoft Designer) the first CBOR "name" key precedes a binary hash field, not the generator string, so _cbor_text_after returns control-char garbage. Guard with isprintable() to drop it; issuer detection (byte-search) and the SynthID verdict are unaffected. Adds TestParseChunkGuards covering kept-vs-dropped cases. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-24 15:55:07 -07:00
test-userandClaude Opus 4.7	da0edcbddc	chore(corpus): grow SynthID reference set + document autonomous Chrome collection Adds content positives (OpenAI gpt-image: forest, fisherman, tokyo; Google gemini: fisherman, mug) and SDXL/non-SynthID negatives to the local corpus manifest. Now spans 4 resolutions across 2 vendors (was solid-black only). README: documents driving generation via Chrome MCP -- Gemini single-click download; ChatGPT via in-page fetch+blob (preserves original C2PA bytes, unlike the flaky UI download / a canvas re-encode). Images stay gitignored; only the manifest (sha256 + labels + extracted metadata) and protocol are tracked. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-24 12:46:46 -07:00
test-userandClaude Opus 4.7	c006f9b8b4	docs(roadmap): record next steps for SynthID detector work Captures the forward plan so a future session picks it up: local pixel detector is blocked pending a generation API or raw watermarked dataset (spectral methods shown insufficient); grow the oracle-labeled corpus; replace synthetic non-PNG C2PA fixtures with real ones; and the maintenance debt (idna bump, strict-pyright cleanup) needed for a green maintain.sh. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-24 11:34:43 -07:00
test-userandClaude Opus 4.7	f07ce10c72	feat(metadata): SynthID-source detection, C2PA parser consolidation, corpus + tests Detect SynthID-bearing images via their C2PA companion: a manifest signed by a SynthID-using vendor (Google/OpenAI) on AI-generated content implies an invisible SynthID pixel watermark. Verified end-to-end against the vendor oracles (openai.com/verify, Gemini "Verify with SynthID"). - metadata: synthid_source() + synthid_watermark verdict in get_ai_metadata, surfaced as a `metadata --check` callout. Format-agnostic (PNG caBX parser + JPEG/WebP/AVIF/HEIF/JXL binary scan). - constants: SYNTHID_C2PA_ISSUERS {Google, OpenAI}; +opened/placed actions. - c2pa: single CBOR-aware parser (_cbor_text_after) replaces glitchy regex (fixes fGPT-4o claim_generator); removed duplicate _scan_png_c2pa_chunk from metadata; shared synthid_verdict / synthid_vendors_in helpers. - corpus: scripts/synthid_corpus.py ingest tool + data/synthid_corpus/ (manifest tracked, images gitignored) for a labeled reference set. - tests: +38 across C2PA parser internals, extract/inject round-trip, ISOBMFF container stripping, all IPTC AI markers, and invisible watermark strength tiers (SynthID/StableSignature/TreeRing/StegaStamp/RingID/RivaGAN/...). Pixel-level SynthID detection remains out of reach locally (Google's decoder is proprietary); a from-scratch spectral pilot confirmed it does not separate real content. See CLAUDE.md for the full evaluation. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-24 11:32:46 -07:00
test-userandClaude Opus 4.7	c1ff4e1cd9	CLAUDE.md: document maintain.sh in Test and lint section Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-23 15:02:06 -07:00
test-userandClaude Opus 4.7	95606ddd5d	docs: SynthID v2 defeat by SDXL pipeline now verified end-to-end locally Local SDXL run on a Gemini 3 Pro output (snowboard scene, 2816x1536), seed 42, strength 0.05, steps 50, ~10 min on MPS. Gemini app's "Verify with SynthID" returned "no SynthID watermark detected" on the cleaned file. This closes the verification gap noted in v0.4.0 release notes and confirms architectural equivalence to the raiw-app production fal-ai/fast-sdxl path. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-17 18:12:56 -07:00
test-userandClaude Opus 4.7	578e229713	style(cli): fix closing paren indentation in cmd_batch Whitespace-only ruff format alignment, no functional change. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	2026-05-17 13:58:35 -07:00