Finalize video package boundary and product wording

.gitignore

@@ -4,6 +4,8 @@ __pycache__/
 *.egg-info/
 dist/
 build/
+tmp/
+.sc/
 
 # Environment secrets
 .env

CLAUDE.md

@@ -38,7 +38,7 @@ Per-command exit-code semantics (the no-signal / GPU-missing skip branches), tes
 
 - **CI** (`.github/workflows/test.yml`): runs on push to `main` + every PR. A `lint` job (ubuntu: `ruff check` + `ruff format --check`) plus a `test` matrix (ubuntu/macos/windows x py3.10/3.12) that does `uv sync --frozen --extra dev` then `pytest`. The matrix installs only core + dev (no `gpu` extra), so the GPU/model-running tests skip there and it exercises the metadata/identify/visible/cv2-eraser surface on all three OSes. Keep `uv.lock` valid (don't break `--frozen`) when editing `pyproject.toml`.
 - Dependency PR checks run the GitHub merge ref against current `main`, not the contributor branch in isolation. If `main` moves after the dependency branch was opened, merge current `main` locally and rerun the full gate; a new linter can expose stale directives in code that landed later.
-- **Release flow + distribution channels** (PyPI publish via `publish.yml`/`uv publish`, the automated Homebrew-tap + HF-Space bumps in `distribute.yml`, conda-forge, ComfyUI Registry, the sdist `data/` exclusion, hatchling pin history): see `docs/release-and-distribution.md` before cutting a release.
+- **Release flow + distribution channels** (PyPI publish via `publish.yml`/`uv publish`, the automated Homebrew-tap + HF-Space bumps in `distribute.yml`, conda-forge, ComfyUI Registry, the sdist public allowlist, hatchling pin history): see `docs/release-and-distribution.md` before cutting a release.
 - `bash maintain.sh` — uv-outdated, uv-secure, ruff check/fix, ruff format, pyright (scoped `src/`, see the OOM note below), pytest -n auto. The helper tools live in the `dev` extra (`pytest-xdist`, plus `uv-outdated`/`uv-secure` marker-gated to py3.12+ so the py3.10 resolution stays solvable) — a bare env without `--extra dev` does not have them.
 - **Strict pyright is clean across `src/` (0 errors).** The cv2/torch/diffusers boundary files (`gemini_engine`, `region_eraser`, `doubao_engine`, `humanizer`, `invisible_engine`, `noai/watermark_remover`) carry a documented per-file `# pyright:` relax pragma that turns off only the unknown-type / untyped-third-party rules — those libs ship no usable types, so strict typing there fights the ecosystem. Pure-logic files stay fully strict; `typings/piexif/__init__.pyi` is a local stub so `metadata.py`/`extractor.py` resolve piexif. Public ndarray-returning signatures on the relaxed engines are still annotated `NDArray[Any]` so strict consumers (`cli.py`) stay clean. When touching a relaxed file, prefer fixing real issues over widening the pragma; keep the pragma scoped to genuinely-untyped boundaries. The `uv-secure` CVE-resolution history (idna/aiohttp bumps, retired basicsr, the dismissed torch `GHSA-rrmf-rvhw-rf47`) lives in `docs/release-and-distribution.md` — read it before re-triaging a dependency alert.
 - **Full-project `uv run pyright` (no path) OOMs/crashes node on this ML-heavy repo** (emits a `libnode` stack frame, no summary) — a known environment limit, not a code error. Gate with `uv run --extra dev --extra gpu pyright src/` (completes, authoritative) or scope to changed files; also run `uv run ruff check` and `uv run pytest` directly.

docs/known-limitations.md

@@ -212,10 +212,10 @@ WebM, Matroska, MP3, WAV, FLAC, OGG, Opus, and AAC container metadata is strippe
 through ffmpeg with stream copying. The operation fails if ffmpeg is absent or
 cannot parse the input.
 
-### Video pixel removal is provider-specific and still experimental
+### Video pixel removal is provider-specific
 
-The experimental `video metadata` command and high level video API inspect and
-strip supported AI provenance metadata without transcoding streams.
+The `video metadata` command and high level video API inspect and strip
+supported AI provenance metadata without transcoding streams.
 
 `video visible` and `remove_video_visible` additionally support the moving
 Sora 2 mascot and wordmark, the current Veo four-point diamond, the legacy

docs/release-and-distribution.md

@@ -65,10 +65,14 @@ current line according to the environment.
 
 The wheel includes the package under `src/`.
 
-The source distribution explicitly excludes `/data` through
-`[tool.hatch.build.targets.sdist]` in `pyproject.toml`. Keep that exclusion:
-calibration captures and test corpora do not belong in the published package
-archive.
+The source distribution uses an explicit allowlist for `/src`, `/LICENSE`,
+`/README.md`, and `/pyproject.toml` through
+`[tool.hatch.build.targets.sdist]` in `pyproject.toml`. It also defensively
+excludes `/data`, `/tmp`, and `/.sc`. Keep both controls: calibration captures,
+test corpora, generated research outputs, and local session state do not belong
+in the published package archive. The matching local-root entries in
+`.gitignore` prevent accidental commits, but are not a substitute for the build
+boundary because hatchling may include untracked files.
 
 ## Build backend
 

docs/verification-plan.md

@@ -242,9 +242,9 @@ was negative on both carriers, and a weaker candidate was negative on one. On
 ordinary Gemini to ignore and reinterpret the detector result. Those follow-ups
 were mistakenly treated as a stricter oracle; they were not detector reruns.
 The implementation remains exposed as `video invisible` and
-`remove_video_invisible`. Its default is a calibrated operating point, while
-each important result still requires a fresh control-positive,
-candidate-negative pair.
+`remove_video_invisible`. Its oracle-certified default is the product operating
+point. A fresh control-positive, candidate-negative pair remains an optional
+per-file audit after provider changes or for unusually important files.
 
 The 2026-07-31 full-clip check added a public eight-second Veo carrier. The
 source and the complete `0.10` product output were both detected, proving the

packaging/conda/recipe.yaml

@@ -59,7 +59,7 @@ about:
     Detect and remove registered visible AI-provenance marks and strip
     AI-provenance metadata (C2PA, EXIF, IPTC, and PNG text chunks) from images
     and supported video containers. The core package covers the identify,
-    metadata, visible, erase, and experimental video command surface. Optional
+    metadata, visible, erase, and product-oriented video command surface. Optional
     pip extras add SynthID diffusion removal and additional invisible-watermark
     detectors.
   license: Apache-2.0

pyproject.toml

@@ -191,9 +191,20 @@ build-backend = "hatchling.build"
 packages = ["src/remove_ai_watermarks"]
 
 [tool.hatch.build.targets.sdist]
-# Keep the source distribution small: ship the package and metadata, not
-# calibration or evaluation data under data/. The wheel ships only src/.
-exclude = ["/data"]
+# Keep the source distribution small and public-safe: ship tracked source and
+# metadata, not corpora or local research/session artifacts. The wheel ships
+# only src/.
+include = [
+    "/src",
+    "/LICENSE",
+    "/README.md",
+    "/pyproject.toml",
+]
+exclude = [
+    "/data",
+    "/tmp",
+    "/.sc",
+]
 
 [tool.pytest.ini_options]
 testpaths = ["tests"]

src/remove_ai_watermarks/video_synthid.py

@@ -1,4 +1,4 @@
-"""Shared configuration for experimental video SynthID regeneration."""
+"""Shared configuration for oracle-certified video SynthID removal."""
 
 DEFAULT_VIDEO_SYNTHID_VAE = "stabilityai/sd-vae-ft-mse"
 DEFAULT_VIDEO_SYNTHID_NOISE_STD = 0.15

tests/test_distribution_boundary.py

@@ -0,0 +1,22 @@
+"""Distribution-boundary regression tests."""
+
+from __future__ import annotations
+
+import re
+from pathlib import Path
+
+
+def _array_values(section: str, key: str) -> set[str]:
+    match = re.search(rf"(?ms)^{key}\s*=\s*\[(.*?)^\]", section)
+    assert match is not None
+    return set(re.findall(r'"([^"]+)"', match.group(1)))
+
+
+def test_sdist_has_explicit_public_boundary() -> None:
+    """Hatchling must publish only package source and required metadata."""
+    root = Path(__file__).resolve().parents[1]
+    config = (root / "pyproject.toml").read_text(encoding="utf-8")
+    sdist_config = config.split("[tool.hatch.build.targets.sdist]", maxsplit=1)[1].split("\n[", maxsplit=1)[0]
+
+    assert _array_values(sdist_config, "include") == {"/src", "/LICENSE", "/README.md", "/pyproject.toml"}
+    assert {"/data", "/tmp", "/.sc"} <= _array_values(sdist_config, "exclude")