Files
gstack/bin/gstack-redact-prepush
T
+22 8bb6d65270 v1.91.66.0 fix: /autoplan without permission prompts after background reviewers, 34 contributor PRs from the backlog triage (#3087)
* docs: fix CONTRIBUTING.md test-tiers link in TESTING_INTERNALS

The link resolved relative to docs/, pointing at a nonexistent
docs/CONTRIBUTING.md. Adopts #3070.

Co-authored-by: thearbajonart <334330110+thearbajonart@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* ci: refresh GitHub Actions SHA pins (setup-buildx 4.4.1, build-push 7.4.0, attest 4.2.2, setup-node 7.0.0)

Rewrite of dependabot #2988. Every bumped `uses:` stays pinned to the full
commit SHA (verified against the upstream tags) and now carries its exact
version comment, including the lines that had none. setup-node 6.5.0 -> 7.0.0
is a major bump; its notes list an ESM migration, cache outputs and removal of
a dummy NODE_AUTH_TOKEN export, none of which affect our node-version-only use.
actionlint 1.7.11 (CI image) and 1.7.12 are clean.

Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* docs(openclaw): stop advising a per-spawn env for spawned sessions

OpenClaw's sessions_spawn has no env argument (openclaw/openclaw#51654
was closed without adding one), so the documented
env: { OPENCLAW_SESSION: "1" } silently did nothing. Point users at the
ACP harness or Gateway environment instead and guard the old advice.
Adopts #2749.

Co-authored-by: DanCanadian <211164803+DanCanadian@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* fix(benchmark): scope /benchmark routing to web page performance

/benchmark claimed the bare words "benchmark" and "performance", so model
and eval benchmark requests could route to the web performance skill
instead of /benchmark-models. Narrow the description, Use-when list and
triggers to web-specific phrases, point model/skill benchmarks at
/benchmark-models, and split the root router line the same way. The
catalog lead stays within the discovery budget ("Web performance ...").

Adopts #2302 and closes duplicate #2083 (rewritten on current main).

Co-authored-by: Nader Helmy <31414549+creatornader@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* fix(llms.txt): link skills by generated path, drop host-skipped skills

gen-llms-txt built every skill link as <name>/SKILL.md, so the router
pointed at a nonexistent gstack/SKILL.md (it is the root SKILL.md) and
/claude-code pointed at a SKILL.md the Claude host never generates.
Links now use the template's generated output path, and skills the
Claude host skips are left out of the index. Adopts #2279.

Co-authored-by: Nader Helmy <creator.nader@gmail.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* fix(ci): install Bun from the SHA-256-verified release archive in GitLab CI and Ubicloud setup

Rewrite of #1713. .gitlab-ci.yml's .setup-bun anchor and
scripts/ubicloud/setup-free-suite.sh still piped https://bun.sh/install into
bash. Both now download the bun-v$BUN_VERSION linux-x64 release zip, check it
against the same SHA-256 Dockerfile.ci uses (#1706) before unzipping, install
it to ~/.bun/bin and assert `bun --version`. test/ci-image-cli-pin.test.ts runs
curlPipedToShell over both files and binds their hash and BUN_VERSION to
Dockerfile.ci's amd64 arm and package.json's engines floor, so a Bun bump has
to move all of them together. Verified by running the GitLab anchor in
debian:stable-slim (bash and sh) and the Ubicloud block in a temp HOME.

Co-authored-by: Zee Zheng <228040220+spacegeologist@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* fix(browse): viewport auto unpins a pinned viewport without a restart

`viewport WxH` pinned the page with no way back short of `browse restart`,
which also drops a headed /connect-chrome session. `viewport auto` (aliases
`reset`, `unpin`) now undoes the pin:

- headed: the context already launches with `viewport: null`, but Playwright
  cannot drop a page's emulated size (and re-applies it on new frame
  sessions), so the active tab gets a fresh page in the same context at the
  same URL or load-html content. Cookies and storage persist, the tab id and
  ownership stay, refs reset as on navigation, and the page follows the real
  window again.
- headless: back to the 1280x720 launch default, tracked viewport included.

`auto` with --scale is rejected. Fixes #1059. Adopts #2782 and #1881.

Co-authored-by: Jack Thomas <17882510+loulanyue@users.noreply.github.com>
Co-authored-by: Jayesh Betala <jayesh.betala7@gmail.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* fix(browse): screenshot accepts --clip=x,y,w,h and --selector=<css>

Agent-generated commands often use the GNU `--flag=value` form, which threw
"Unknown screenshot flag". Both value flags now accept it, through the same
validation and conflict checks as the space-separated form. Adopts #1563.

Co-authored-by: techcenter68 <75930234+techcenter68@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* fix(benchmark): price Codex cached input at the cache rate, stop calling Gemini OAuth ready

Codex reports cached_input_tokens as a subset of input_tokens, but the
adapter billed all of input_tokens at the full input rate. Split them
into disjoint uncached and cached counts (what estimateCostUsd expects)
and show new+cached input in the table and Markdown reports. A stored
Gemini OAuth file alone is now reported as not ready, with a reason
naming GEMINI_API_KEY; key-based auth stays ready. Adopts the focused
subset of #2283; the capacity error code and stdin change are left out
(no captured failure or reproduced hang).

Co-authored-by: Mattias Petersson <mattias.petersson@gmail.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* fix(review): catch loosened-input consumers and stale user-facing strings

The Enum & Value Completeness check fired only on newly added values, so a
diff that widens what a validator, parser or guard accepts, and breaks an
unchanged downstream consumer outside the diff, was a blind spot. Widen the
check's trigger (heading unchanged) and add a consumer-grep bullet. Teach
the maintainability specialist to flag user-facing strings whose guarding
condition changed while the text did not.

Adopts #2141 (rewritten on current main; the section heading is kept).

Co-authored-by: Arno V <38716+aversini@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* fix(global-discover): bucket Codex sessions by originator

Every Codex rollout counted as one undifferentiated codex bucket, so
/retro global could not tell interactive Codex Desktop work from
codex_exec runs or Claude Code-driven calls. Normalize
session_meta.payload.originator into desktop, exec, claude_code or
other (missing, unknown and non-string values), report the split per
repo as codex_originators and globally as tools.codex.originators, and
print it in the summary. Existing totals are unchanged. Adopts #1488.
Refs #1315.

Co-authored-by: 0xDevNinja <manmit0x@gmail.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* fix(hook-check): cover skill frontmatter hooks, sourced helpers, conflict markers and the caller's tsconfig

Rewrite of #3066. bin/gstack-hook-check only listed setup's hooks plus
/autoplan's by name, so the /careful and /freeze PreToolUse checks (also run
by /guard and /investigate), /plan-ceo-review's hand-off hook and the helpers
they source were never parse-checked before Claude Code ran them.

- The hook list is setup's _hook_command_path arguments plus every gstack path
  on a `command:` line in a generated */SKILL.md frontmatter (9 hooks, up from
  6). The generated files are read rather than the .tmpl sources because the
  /autoplan and /plan-ceo-review commands come from resolvers.
- Each shell hook's gstack-tree helpers (`.`/`source` lines, assignments and
  shellcheck source= directives, followed transitively) are parsed too:
  careful/bin/hook-extract.sh and bin/gstack-state-root.sh for /careful and
  /freeze.
- The shim, its helpers, each TypeScript entry and every local module the
  bundle used (bun build --metafile; entry only when bun lacks it) are scanned
  for conflict-marker lines, which still parse inside heredocs, strings and
  template literals.
- bun build runs from the gstack root, so the caller's broken tsconfig.json or
  bunfig.toml no longer fails every TypeScript hook. Lists are newline
  separated, so a root with spaces works.

Tests in test/setup-hook-syntax-gate.test.ts cover each case; the gstack-doctor
fixture now copies autoplan/SKILL.md, whose frontmatter registers the hook its
hooks-row test breaks. docs/troubleshooting.md describes the wider check.

Co-authored-by: Benjamin D. Smith <7134745+BenjaminDSmithy@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* fix(review): extend defensive-security framing to the security specialist and Red Team

The "authorized defensive-security review" framing and fixture summary mode
lived only in the adversarial subagent prompt. The security specialist and
the Red Team also reason adversarially over the full diff, so repos whose
test corpus is attack payloads could still get those passes refused.

Move both paragraphs into scripts/resolvers/defensive-review.ts (the
adversarial render stays byte-identical), state them once in the specialist
dispatch for the security specialist and the Red Team, and have the Red
Team prompt start from them. Testing and the other specialists keep full
fixture reads; the codex host still strips Review Army. Parity caps for
ship and review rise by the measured 1,380 bytes each (carve-guards notes).

Adopts #1921 (rewritten on current main).

Co-authored-by: Brian Majewski <1693720+bmajewski@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* fix(team-init): make the required-mode hook shell-neutral (node CommonJS, no shell variables)

Rewrite of #2217. Required team mode wrote a bash check-gstack.sh and
registered it as "$CLAUDE_PROJECT_DIR/.claude/hooks/check-gstack.sh", which
only works when the hook runner is a POSIX shell. Under cmd.exe or PowerShell
(Windows, Copilot CLI) the variable is not expanded and the .sh cannot run, so
an intended allow or deny became a hook error.

- The hook is now .claude/hooks/check-gstack.cjs: the same root search
  (GSTACK_ROOT, every host's global root, the migrated repo), with the home
  directory from HOME, then USERPROFILE, then os.homedir(). It prints {} when
  gstack resolves and the PreToolUse deny JSON with exit 2 when not.
- The registered command is `node -e "require(require('path').resolve(
  process.env.CLAUDE_PROJECT_DIR || '.', '.claude', 'hooks',
  'check-gstack.cjs'))"`: no `$` or `%`, so sh, cmd.exe and PowerShell all
  pass it through unchanged.
- An existing check-gstack registration (matcher Skill or Skill|skill) is
  replaced in place; duplicates are dropped and unrelated hooks in the same
  entry kept. The legacy .sh is deleted only after the settings write
  succeeds (a tracked one joins the suggested git add), and is kept when bun
  is missing or the write fails, which now reports instead of claiming success.

Tests: team-mode and the hook-schema test run the command as registered;
test/team-init-hook-shell-neutral.test.ts runs it through the platform's
default shell (cmd.exe on the Windows lane, which curates it in) and through
PowerShell when installed (verified here in the pwsh 7 container). The
routing-probe check for the bash `_GSTACK_ROOT` gate now pins the same
contract in the .cjs: every host root probed, block only when none resolves.

Co-authored-by: neallee2012 <1499497+neallee2012@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* fix(setup-gbrain): register the gbrain MCP through bun.exe on Windows so no console window opens

Rewrite of #2007, folding #1963's MCP notes. On Windows the local stdio
registration stored the bun `gbrain` binstub; gbrain.exe starts bun in a new
console, so every Claude Code session opened a visible terminal for the MCP
(the contributor's process trees show it; bun.exe launched directly attaches
to Claude Code's headless ConPTY).

Under MINGW/MSYS/Cygwin the skill now resolves BUN_ROOT from BUN_INSTALL or
~/.bun and registers `$BUN_ROOT/bin/bun.exe <global node_modules
gbrain/src/cli.ts> serve`. That path also covers a `bun link`ed clone, whose
global node_modules entry links to it; gstack-gbrain-install records no other
clone location. When bun.exe or the entry is missing it registers
`$BUN_ROOT/bin/gbrain.exe` by its full name, because Claude Code does not
apply PATHEXT to a stored path (#1963). POSIX keeps the binstub line. The
skill also notes that `claude mcp list` can show "Failed to connect" while a
live session holds a local PGLite brain's single-writer lock (#1963).

test/setup-gbrain-windows-mcp.test.ts runs the generated fence under bash with
stub uname/claude and a temp HOME for MINGW64, MSYS with BUN_INSTALL, a
missing entry, and Linux. It is listed as Windows-incompatible (POSIX stubs);
the console behavior needs a manual Windows check.

Co-authored-by: Max Hogan <max@maxhogan.dev>
Co-authored-by: maxpetrusenkoagent <280464918+maxpetrusenkoagent@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* fix(hosts): rewrite every "Agent tool" phrasing for agent-runtime hosts

The Agent-tool rewrite for hermes, gbrain, openclaw and factory keyed only
on "use the Agent tool", but templates mostly say "via the Agent tool",
"with the Agent tool", "multiple Agent tool calls" or "Use the Agent tool",
so the Claude tool name leaked into those hosts' skills and sections. Add
ordered rules after the existing key (possessive, "the Agent tool", then
bare "Agent tool") mapping to sessions_spawn (gbrain/openclaw),
delegate_task (hermes) and delegation (factory). No goldens change.

Adopts #1935 (rewritten on current main, without the new goldens).

Co-authored-by: katlun-lgtm <264247399+katlun-lgtm@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* fix(setup-deploy): Vercel auto-deploys only when connected to the Git repo

/setup-deploy stated that Vercel always deploys on push, which is false for
CLI-only projects with no Git integration, and its nothing-detected path did
not say why such a project is invisible to detection. Qualify the Vercel
trigger (ask for and record the CLI deploy command otherwise) and note that
`vercel link` creates .vercel/ so a re-run detects the project.

Adopts the Vercel accuracy part of #2992 (rewritten on current main); the
third-party bring-up tool pointer is not adopted.

Co-authored-by: mikehasa <91774998+mikehasa@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* fix(retro): persist recommendations so the next retro detects follow-through

/retro produced "3 Things to Improve" but never saved them in its JSON
snapshot, so the next retro could not tell whether last week's
recommendations were acted on. Step 13 now drafts the items before saving
and always writes a `recommendations` array ({category, text}, [] when
empty); Step 12 scores each prior recommendation addressed/partial/open
against this window's commits, files and metrics and renders a
follow-through line, skipping silently for snapshots without the key. The
report format ties the items to the saved array. Global mode is unchanged.

Adopts #1853 (rewritten on current main). Refs #1834.

Co-authored-by: Jayesh Betala <jayesh.betala7@gmail.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* fix(global-discover): count the interactive Codex CLI (codex_cli_rs) in its own cli bucket

* feat(developer-profile): --reconcile backfills office-hours tenure from timeline.jsonl

sessions[] is written only at office-hours Phase 4.5, so runs that ended
earlier never counted toward tenure and the tier undercounted. Add a manual
`--reconcile` that scans projects/*/timeline.jsonl for office-hours
`completed` events and appends a count-only row ({date, mode, project_slug,
signal_count: 0, signals: [], backfilled: true}) for each run with no
sessions[] entry for the same project within 60 minutes, extended back by a
duration_s under 8h. It is idempotent, skips malformed lines, and leaves the
file untouched on a no-op. `--read` counts backfilled rows for SESSION_COUNT
and TIER but takes LAST_*, DESIGN_*, CROSS_PROJECT and the nudge from logged
sessions only. Not wired into the preamble yet; the office-hours greeting
guard for backfill-only profiles is left to the office-hours owner.

Adopts #2668 (rewritten on current main). Closes part (b) of #2657.

Co-authored-by: Briac <137724549+briacSck@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* fix(autoplan): the publication hook approves /autoplan's own Reads after the skill turn

Claude Code applies a skill's allowed-tools only to the turn that invoked it.
On 2.1.292 the default fork-subagent mode runs reviewers in the background, and
the completion notice starts a new turn, so later Reads outside the working
directory (the next phase's section file, the close packet) raised permission
cards. Skill hooks stay registered for the session, and a PreToolUse "allow"
skips the card (reproduced in a PTY on 2.1.292).

autoplan/bin/owned-read.ts approves a parent Read only when the guard itself
returned no output, and only for this installation's autoplan/sections/*.md
(canonical path or the gbrain :user render) or an immutable phase artifact in
<repo>/.gstack/tmp/autoplan/autoplan-<phase>-*/ whose repo is the session's
project directory or contains it. Denials, unverified allows, child calls and
every other path are unchanged. renderSectionBase and userRenderRoot move
there with it.

Tests that asserted a verified driver Read as {} now expect the approval;
autoplan-owned-read covers the shipped shim, the close packet without a
journal, verified artifacts through the guard, look-alikes (foreign install,
writable copy, symlink alias, other repository, other names) and a guard
denial staying a denial.

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* fix(autoplan): restore point and phase artifacts live in the project's .gstack/tmp/autoplan/

Skill hooks do not run inside subagents, so a reviewer launched after a
background completion notice raised a card for its own input (native-prompt.md
under the state root). Step 1 now puts RESTORE_PATH in the repository's
git-excluded .gstack/tmp/autoplan/ (added to .git/info/exclude, mode 700), and
the snapshot tool writes every phase artifact beside it, so /autoplan and its
reviewers read them from the working directory without a card. The guard's
bindings are unchanged: they anchor on the restore point's directory. Step 1
still prints the slug, which the old RESTORE_PATH carried.

A free test runs the generated Step 1 block twice from a repository
subdirectory: the restore point lands in <repo>/.gstack/tmp/autoplan/, the
exclude line is written once and git status stays clean. The autoplan parity
cap rises by the measured bytes (115,270 / 101,979 = 1.1303).

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* test(evals): autoplan-guard-pty fails on any permission card after the reviewer's notice turn

The scripted boundary now mirrors Step 1's layout (restore point and artifacts
in the project's .gstack/tmp/autoplan/) and continues past the Phase 2 entry to
dispatch the Phase 2 reviewer after the CEO reviewer's completion notice. The
case asserts four verified guard allows, that the close packet Read, the Phase 2
section Read and the Phase 2 dispatch came after a task-notification turn, that
the Phase 2 reviewer read its input, and that no permission card appeared at
all. Still one trial on haiku, about 75-85 seconds.

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* docs(autoplan): permission cards after a background reviewer, and where /autoplan keeps its files

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* fix(outside-review): read a trailing '— Medium.' or '(High)' severity on a finding line, the shape Codex now writes

* feat(design): honor OPENAI_BASE_URL for every design API call

One resolver, openaiUrl() in design/src/auth.ts, now builds every OpenAI
URL the design binary, its prototype and its live model check use. Setting
OPENAI_BASE_URL sends image generation and screenshot analysis to an
OpenAI-compatible gateway; unset or blank keeps api.openai.com. The
prototype script now goes through receiptedFetch, so its egress is
receipted like the rest, and receipts already record the gateway host.

A source scan keeps any design file outside auth.ts from hardcoding an
OpenAI API URL again, and a request-shape test proves the gateway reaches
both the vision and image endpoints.

Co-authored-by: JackJin <1037461232@qq.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* review: testing specialist treats negative assertions as decisions

A new assertion that something is absent records a behavior removal and
needs cited intent. A rename or removal in the diff can empty a standing
negative assertion so it passes forever; the specialist now greps the suite
for negative assertions naming the old value.

Adopted from the testing-specialist part of a community PR; the scope-check
hunk targeted text that no longer exists.

Co-authored-by: EmilianoU26 <314541067+EmilianoU26@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* review: security specialist checks authorization a route-guard read misses

Four access-control classes where the guard is correct and the object still
leaks: error pages composed outside the guard, ownership surviving a role
downgrade, list/detail/download paths checked as one, and a session mutated
before the identity-switch callback is validated.

Adopted to review/specialists/security.md only; the /cso hunk targeted a
bullet list that is now an OWASP/ASVS table.

Co-authored-by: EmilianoU26 <314541067+EmilianoU26@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* review: zero findings is valid; suppress six common false positives

The checklist now says a clean diff may come back clean, and suppresses
framework-owned error handling, long exhaustive switches, null derefs a
type guard already narrows, N+1 on fixed-cardinality loops, hardcoded test
fixture values and non-cryptographic Math.random().

The contributor's "missing input validation on an internal function" and
"missing await on fire-and-forget" suppressions are left out: both can hide
real bugs.

Co-authored-by: frosimanuel <frosimanuel@gmail.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* docs(todos): parked contributor ideas by theme; /autoplan permission follow-ups

* design-review: capture code tokens before universal principles

/design-review read only DESIGN.md and otherwise fell back to universal
principles, so a project whose system lives in CSS custom properties,
tailwind.config or a theme file got audited and mocked up against an
invented aesthetic. The skill now reads code tokens before falling back,
and the target-mockup brief reuses the captured colors, fonts, radii and
spacing, changing layout only.

The contributor's optional microcopy checks (domain vocabulary, human
register) are not adopted here.

Co-authored-by: Tom Lam <tomlam0811@berkeley.edu>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* feat(sidebar): add a terminal text-size picker to the toolbar

The sidebar terminal toolbar gets an XS-XL text-size picker beside
Restart. SM (13px) stays the default, so nothing changes until a user
picks a size; the choice persists in chrome.storage.local and applies
to the live terminal, re-fitting xterm and re-sending the PTY size two
frames later so the new character cell is measured first.

A VM test runs the real sidepanel-terminal.js through one auto-connect
and checks the creation size, the stored-size fallback, persistence and
the live resize.

Co-authored-by: Yotam <yotsurfing@gmail.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* feat(design-shotgun): generate with Codex's built-in $imagegen on Codex

On the Codex host, /design-shotgun now generates and edits mockups with
Codex's documented $imagegen skill in its default built-in mode (the
image_gen tool), so users signed in with ChatGPT need no OPENAI_API_KEY.
The design binary stays for the local comparison board only, images are
published with gstack-design-claim as before, and screenshot evolution
loads the captured page with view_image before an $imagegen edit.

Step 3c-3d moved into a DESIGN_SHOTGUN_GENERATION resolver that returns
the existing design-binary text unchanged for every other host; the
Claude output differs only in two host-neutral rewordings of the
evolve-from-screenshot pointers. DESIGN_SETUP and the regenerate step of
the board loop gain matching Codex-only text for design-shotgun.

Co-authored-by: Jinzhe <hwang.phy@gmail.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* voice: ban the "load-bearing" stock phrase

Models overuse "load-bearing"; it joins the AI-vocabulary ban line in both
voice tiers. The tier-2 voice directive is also tightened (the meta
"compressed for runtime" tag and the bounded closer's restated exemption
clause) so the shared preamble shrinks by 163 bytes per tier-2+ skill,
which also pays for the reply-language rule that follows. The ship goldens
are refreshed for the new render.

Co-authored-by: Nader Helmy <creator.nader@gmail.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* voice: reply in the language of the user's latest message

Every skill's voice directive now tells the agent to reply in the language
of the user's latest message unless asked otherwise, so long English skill
text no longer drags a Chinese (or any non-English) session into English.
Mixed-language sessions follow the latest message. Code, commands, paths,
identifiers and quoted output stay verbatim, and at tier 2+ so do the
machine-read question markers (D<N>, option letters, (recommended)).

The rule costs 205 bytes per tier-2+ skill and 144 per tier-1 skill. A
further 25-byte cut to the bounded closer's exemption list leaves the net
for this wave at +17 bytes per tier-2+ skill and +158 per tier-1 skill.
review, qa and browse caps rise with measured notes; the ship goldens are
refreshed.

Co-authored-by: Xianying Tan <shrektan@126.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* feat(redact-prepush): committed exact-span allowlist at .gstack-redact-allowlist

The pre-push hook never fed the engine's allowlist option, so a reviewed,
known-benign value (a public dev-compose URL, a documented sample key) warned
or blocked every push, and the only ways out were a bypass or reshaping code
until the detector stopped matching.

The hook now reads the repo-root .gstack-redact-allowlist from each pushed
commit (never the working tree), one exact matched span per line with #
comments. Only a finding whose matched text equals an entry is suppressed: a password alone, a
substring, or a different key still reports and still blocks, and engine.*
unscanned findings are untouched. Every push that carries entries prints the
entry count and each suppressed finding's rule and file line. Oversized files
(over 64 KiB) are ignored. gstack.redact.allowEmail keeps working as before.

Co-authored-by: Boyard Engels <boyard.engels@gmail.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

* feat(brain-sync): opt-in removal publishing with a bulk-removal valve

gstack-brain-sync never published a removal: a deleted synced file, or the
old directory after a project slug migration, stayed on the remote forever
and gbrain indexed both copies.

New setting artifacts_sync_removals (off by default). When on, each drain
commits and pushes, receipted like every push, the removal of tracked files
that are gone from disk and match the sync allowlist. Transcript pages are
never removed here, and a decision log goes only together with its active
snapshot. One valve: a drain that would remove more than 20 files at once
publishes none and says so. `gstack-brain-sync --publish-removals` lists the
pending removals and `--publish-removals --yes` publishes them on demand,
whatever the setting.

Co-authored-by: Jey2311 <5177215+Jey2311@users.noreply.github.com>

* feat(windows): setup.ps1 checks prerequisites and runs ./setup in Git Bash

A PowerShell entry point for Windows users. It checks that Git for Windows,
Bun and Node.js are on PATH (printing the winget command for each missing one
and changing nothing), finds Git Bash from git.exe rather than a PATH bash
that may be WSL's, and runs ./setup from the checkout with every argument
passed as its own argv entry to a constant bash script. No argument is
interpolated into shell source. ./setup itself is unchanged.

Co-authored-by: Jim McKeeth <jim@mckeeth.org>

* v1.91.63.0: CHANGELOG, version

* fix(ci): link bunx after the SHA-verified Bun install in GitLab CI and the Ubicloud setup

* fix(redact-prepush): strip every parent column from combined-diff lines, so a merged release bump is exempt like any other

* fix(hook-check): treat a Windows drive path as absolute (cygpath when present), so --list finds the skill hooks from a native caller

* fix(windows): owned-read requires the read-only attribute on Windows; setup.ps1 passes arguments NUL-separated through a private file; Windows-only tests expect the owned-read approval

* perf(hook-check): bundle every hook's TypeScript entries in one bun process instead of two bun starts per entry (seconds each on Windows)

* perf(hook-check): read all skill frontmatter in one awk and scan conflict markers with one grep per hook (half the process starts)

---------

Co-authored-by: thearbajonart <334330110+thearbajonart@users.noreply.github.com>
Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: DanCanadian <211164803+DanCanadian@users.noreply.github.com>
Co-authored-by: Nader Helmy <31414549+creatornader@users.noreply.github.com>
Co-authored-by: Nader Helmy <creator.nader@gmail.com>
Co-authored-by: Zee Zheng <228040220+spacegeologist@users.noreply.github.com>
Co-authored-by: Jack Thomas <17882510+loulanyue@users.noreply.github.com>
Co-authored-by: Jayesh Betala <jayesh.betala7@gmail.com>
Co-authored-by: techcenter68 <75930234+techcenter68@users.noreply.github.com>
Co-authored-by: Mattias Petersson <mattias.petersson@gmail.com>
Co-authored-by: Arno V <38716+aversini@users.noreply.github.com>
Co-authored-by: 0xDevNinja <manmit0x@gmail.com>
Co-authored-by: Benjamin D. Smith <7134745+BenjaminDSmithy@users.noreply.github.com>
Co-authored-by: Brian Majewski <1693720+bmajewski@users.noreply.github.com>
Co-authored-by: neallee2012 <1499497+neallee2012@users.noreply.github.com>
Co-authored-by: Max Hogan <max@maxhogan.dev>
Co-authored-by: maxpetrusenkoagent <280464918+maxpetrusenkoagent@users.noreply.github.com>
Co-authored-by: katlun-lgtm <264247399+katlun-lgtm@users.noreply.github.com>
Co-authored-by: mikehasa <91774998+mikehasa@users.noreply.github.com>
Co-authored-by: Briac <137724549+briacSck@users.noreply.github.com>
Co-authored-by: JackJin <1037461232@qq.com>
Co-authored-by: EmilianoU26 <314541067+EmilianoU26@users.noreply.github.com>
Co-authored-by: frosimanuel <frosimanuel@gmail.com>
Co-authored-by: Tom Lam <tomlam0811@berkeley.edu>
Co-authored-by: Yotam <yotsurfing@gmail.com>
Co-authored-by: Jinzhe <hwang.phy@gmail.com>
Co-authored-by: Xianying Tan <shrektan@126.com>
Co-authored-by: Boyard Engels <boyard.engels@gmail.com>
Co-authored-by: Jey2311 <5177215+Jey2311@users.noreply.github.com>
Co-authored-by: Jim McKeeth <jim@mckeeth.org>
2026-10-08 17:40:06 +00:00

764 lines
35 KiB
TypeScript
Executable File

#!/usr/bin/env bun
/**
* gstack-redact-prepush — git pre-push hook that scans the diff being pushed for
* HIGH-severity credentials and blocks the push on a hit.
*
* THIS IS A GUARDRAIL, NOT ENFORCEMENT. `git push --no-verify` bypasses it, as
* does `GSTACK_REDACT_PREPUSH=skip`. It catches accidental credential pushes,
* the most common real-world leak. It does NOT scan history, binary/LFS/submodule
* files, or non-added lines. History scanning is /cso's job.
*
* Git pre-push interface: refs are read from STDIN, one per line:
* <local ref> <local sha> <remote ref> <remote sha>
* We scan the ADDED lines of <remote sha>..<local sha> per ref (what's being
* pushed). Special cases:
* - remote sha all-zeroes → new branch: diff against merge-base with the
* remote's default branch (fallback: scan all commits unique to local ref).
* - local sha all-zeroes → branch delete: nothing to scan, skip.
* - force-push → remote..local still gives the net new content.
*
* Behavior:
* - HIGH finding in added lines → print + exit 1 (block), for public AND private.
* - MEDIUM → warn (non-blocking), one line per finding naming its rule,
* file line and fix, never the matched value. LOW/WARN → silent.
* - `pii.email` skips the pusher's own address, `gstack.redact.allowEmail`
* entries, and addresses already in the destination's commit metadata
* (see emailScanOptions).
* - `.gstack-redact-allowlist`, read from each pushed commit (never the
* working tree), lists exact matched spans reviewed as benign, one per
* line, `#` comments. Only a finding whose matched text equals an entry is
* suppressed; every push that carries entries lists what they suppressed.
* - GSTACK_REDACT_PREPUSH=skip → log + exit 0 (escape valve).
*
* Installed/uninstalled via `gstack-redact install-prepush-hook` (see the
* gstack-redact CLI), which chains any pre-existing hook.
*/
import { spawnSync } from "child_process";
import * as fs from "fs";
import * as path from "path";
import { normalizeWithMap, scan, type Finding, type ScanOptions } from "../lib/redact-engine";
import { mkdirpSync } from "../lib/fs-utils";
import { resolveStateRoot } from "../lib/state-root";
const ZERO = /^0+$/;
let emptyTree: string | undefined;
function emptyTreeOid(): string {
if (emptyTree) return emptyTree;
const oid = gitStrict(["hash-object", "-w", "-t", "tree", "--stdin"]).trim();
if (!/^(?:[0-9a-f]{40}|[0-9a-f]{64})$/.test(oid)) throw new Error("git could not resolve the empty tree");
emptyTree = oid;
return oid;
}
/**
* Permissive git for legitimately-fallible PROBES (symbolic-ref, rev-parse,
* merge-base) where a non-zero exit is normal control flow. The DIFF call
* must NOT use this — see gitStrict (#1946 fail-closed).
*/
function git(args: string[]): string {
const r = spawnSync("git", args, { encoding: "utf8", maxBuffer: 64 * 1024 * 1024 });
return r.status === 0 ? (r.stdout ?? "") : "";
}
/**
* Fail-closed git for the diff that decides whether the push is scanned
* (#1946). status !== 0 covers repo errors; status === null covers a killed
* process AND maxBuffer overflow — the oversized-diff case is exactly where
* a large secret-bearing blob is most likely, so "couldn't read the diff"
* must block, not silently allow.
*/
function gitStrict(args: string[]): string {
const r = spawnSync("git", args, { encoding: "utf8", maxBuffer: 64 * 1024 * 1024 });
// status !== 0 covers BOTH a non-zero exit AND null (process killed by a
// signal or maxBuffer overflow — null !== 0 is true).
if (r.status !== 0) {
throw new Error(
`git ${args[0]} failed (status=${r.status ?? "killed/overflow"}): ${(r.stderr ?? "").slice(0, 300)}`,
);
}
return r.stdout ?? "";
}
/** True when the object exists in the local odb (cat-file -e signals via exit code). */
function objectExists(sha: string): boolean {
const r = spawnSync("git", ["cat-file", "-e", sha], { encoding: "utf8" });
return r.status === 0;
}
/**
* The remote-tracking exclusion used when narrowing to "commits new to the
* remote" (#2592 catch-up merges, #2573 rebased force-pushes).
*
* Narrowed to the PUSH TARGET's namespace (S1): a bare `--remotes` excludes
* commits reachable from ANY remote-tracking ref, so a secret that had only
* ever been fetched from (or pushed to) a private/local-path remote was never
* scanned when later pushed to a PUBLIC remote — "already left this machine"
* is not "already reached THIS remote". Git hands pre-push the push remote's
* name as $1 (and its URL as $2); the installed hook wrapper forwards "$@".
* A URL push has no tracking namespace, even if origin happens to contain
* the same commits. Direct invocations without Git's argv keep the historical
* all-remotes behavior for compatibility.
*/
type PushTarget = { kind: "remote"; name: string } | { kind: "url" } | { kind: "unknown" };
let cachedPushTarget: PushTarget | undefined;
function pushTarget(): PushTarget {
if (cachedPushTarget) return cachedPushTarget;
const name = process.argv[2];
const url = process.argv[3];
if (!name && !url) return cachedPushTarget = { kind: "unknown" };
if (!name || !url) return cachedPushTarget = { kind: "url" };
const remotes = git(["remote"]).split("\n");
if (!remotes.includes(name)) return cachedPushTarget = { kind: "url" };
const pushUrls = git(["remote", "get-url", "--push", "--all", name]).trim().split("\n");
const fetchUrls = git(["remote", "get-url", "--all", name]).trim().split("\n");
return cachedPushTarget = pushUrls.includes(url) && fetchUrls[0] === url
? { kind: "remote", name }
: { kind: "url" };
}
function remotesExclusionArgs(): string[] {
const target = pushTarget();
if (target.kind === "remote") return [`--remotes=${target.name}/*`];
return target.kind === "unknown" ? ["--remotes"] : [];
}
function defaultRemoteBranch(): string | null {
const target = pushTarget();
if (target.kind === "url") return null;
const remote = target.kind === "remote" ? target.name : "origin";
const sym = git(["symbolic-ref", `refs/remotes/${remote}/HEAD`]).trim();
if (sym.startsWith(`refs/remotes/${remote}/`) && git(["rev-parse", "--verify", `${sym}^{commit}`]).trim()) return sym;
for (const b of [`${remote}/main`, `${remote}/master`]) {
const qualified = `refs/remotes/${b}`;
if (git(["rev-parse", "--verify", `${qualified}^{commit}`]).trim()) return qualified;
}
return null;
}
/**
* Base commit for a new remote ref, ordered from most precise to most
* conservative. Returns null when nothing can anchor the range.
*/
function unknownRemoteTipBase(localSha: string): string | null {
// 1. The common case: a merge-base with the remote's default branch.
const defaultBranch = defaultRemoteBranch();
const base = defaultBranch ? git(["merge-base", localSha, defaultBranch]).trim() : "";
if (base) return base;
// 2. No merge-base. The target's default branch may be unavailable
// (named trunk/develop, remote/HEAD unset), or history may be disjoint.
// Anything reachable from localSha but from NO remote-tracking
// branch is what this push actually adds; the parent of its oldest commit
// is the real base.
//
// Without this we drop straight to EMPTY_TREE and re-scan content that is
// already on the remote. That is not merely wasteful, it is wrong in two
// ways: a secret pushed long ago gets re-reported as if THIS push
// introduced it (telling the operator to rotate a key over someone else's
// old commit), and on any real repository the input overshoots the
// engine's byte cap, so `engine.input_too_large` blocks having scanned
// NOTHING — "scans more, never less" inverted into "scans nothing".
//
// The exclusion is scoped to the PUSH TARGET's tracking refs (see
// remotesExclusionArgs): content on some OTHER remote has left this machine,
// but it has not reached the remote being pushed to — a secret that only
// ever hit a private remote must still be scanned on its way to a public
// one (S1).
const exclusions = remotesExclusionArgs();
const newCommits = git(["rev-list", "--reverse", localSha, ...(exclusions.length ? ["--not", ...exclusions] : [])]).trim();
if (newCommits) {
const oldest = newCommits.split("\n")[0];
const parent = git(["rev-parse", "--verify", `${oldest}^`]).trim();
if (parent) return parent;
// Oldest new commit is a root commit: there is no parent to anchor on.
}
// 3. Nothing to anchor on — a genuinely fresh repository with no remote refs.
// Every commit IS new content, so scanning it all is the correct answer.
return null;
}
/**
* The commits this push actually adds — reachable from localSha and from NO
* remote-tracking ref.
*
* ⚠ WHY THIS EXISTS RATHER THAN A TWO-DOT RANGE.
*
* `remoteSha..localSha` is "everything new on this branch", which is NOT the
* same as "everything new to the remote". Merge origin/main into a feature
* branch and every commit main gained since the branch's last push becomes an
* added line — content that is already published, already scanned, and not
* this push's doing.
*
* Two things follow, and both were observed:
*
* · FALSE HIGH FINDINGS. A placeholder connection string in a test fixture,
* already merged to main by someone else, blocked an unrelated push as
* `db.url_with_password` — telling the operator to rotate a credential
* over a fixture they had never touched. A
* guard that cries wolf on catch-up merges is a guard people learn to
* bypass reflexively — which is exactly how a real secret gets through.
* · OVERSIZED SCANS. The comment on SCAN_CHUNK_BYTES below records a
* 1,146,782-byte diff from "a feature branch catching up to a busy main"
* blowing the engine's 1 MiB cap. Same root cause, treated there as a size
* problem and solved by slicing. Narrowing the range fixes the size too.
*
* A two-dot range cannot express this: after merging main, neither the remote
* tip nor the merge-base with main is an ancestor of the other, so no single
* base excludes both. `rev-list --not --remotes=<push-remote>/*` is the
* operation that does, and this file already reasons that way in
* `unknownRemoteTipBase` step 2. The exclusion is scoped to the push target's
* tracking namespace (see remotesExclusionArgs): the upstream commits a catch-up
* merge brings in came from the SAME remote being pushed to, so scoping keeps
* the #2592 fix intact while a secret known only to some OTHER (private)
* remote is still scanned on its way to this one (S1).
*
* Each commit is diffed alone. `--cc` on a merge shows only the conflict
* RESOLUTION — content that exists in no parent — so a secret introduced while
* resolving a merge is still caught, while an ordinary merge contributes
* nothing. Returns null when the notion does not apply, so callers fall back.
*/
function addedLinesFromNewCommits(localSha: string, remoteSha: string): AddedFile[] | null {
// remoteSha is what git TELLS us the remote has, and it is authoritative in a
// way `--remotes` is not: remote-tracking refs can be absent (a fresh clone
// that never fetched, a push to a remote with no tracking ref) or stale. Drop
// it and a repo with no tracking refs excludes NOTHING — every commit ever
// made reads as "new", which re-introduces the false positives from the other
// direction. So it stays the base; `--remotes` only ADDS exclusions on top.
if (ZERO.test(remoteSha) || !objectExists(remoteSha)) return null;
const narrowed = git(["rev-list", localSha, "--not", remoteSha, ...remotesExclusionArgs()]).trim();
if (!narrowed) return null;
// If excluding remote-tracking refs changes nothing, this push has no
// catch-up commits and the plain range already describes it exactly. Defer to
// it. That is not just an optimization: it keeps every push that ISN'T a
// catch-up merge on the original gitStrict diff path, so the fail-closed
// guarantee (#1946) and its regression test keep exercising the code they
// were written for. A narrowing that silently retired that test would be a
// worse trade than the false positives it set out to fix.
const plain = git(["rev-list", `${remoteSha}..${localSha}`]).trim();
const asSet = (s: string) => s.split("\n").filter(Boolean).sort().join("\n");
if (asSet(narrowed) === asSet(plain)) return null;
const shas = narrowed.split("\n").filter(Boolean);
// A rewrite of long history should fall back rather than shell out per commit.
if (shas.length > 500) return null;
const out: AddedFile[] = [];
for (const sha of shas) {
// gitStrict: a failed diff must never read as "nothing added" (#1946).
// Collected per commit so a finding names the commit whose file line it is.
out.push(...collectAddedLines(gitStrict([
"show", "--unified=0", "--no-color", "--no-ext-diff", "--no-textconv",
"--src-prefix=a/", "--dst-prefix=b/", "--cc", "--format=", sha,
]), sha));
}
return out;
}
/**
* Added lines grouped by the file they were added to (`+++ b/` path; empty
* when unknown). `lines[i]` is the file line number of the i-th line of
* `text`, read from the hunk headers; `commit` is set when the lines came from
* one commit's diff rather than the net diff of the pushed range.
*/
interface AddedFile { file: string; text: string; lines: number[]; commit?: string }
/** Return the added lines for a ref update being pushed. */
function addedLinesFor(localSha: string, remoteSha: string): AddedFile[] {
// Preferred ONLY when this push carries catch-up commits: scanning them again
// is the bug. Every other shape falls through to the range logic below.
const fromNew = addedLinesFromNewCommits(localSha, remoteSha);
if (fromNew !== null) return fromNew;
let range: string;
if (ZERO.test(remoteSha) || !objectExists(remoteSha)) {
const base = ZERO.test(remoteSha) ? unknownRemoteTipBase(localSha) : null;
range = base ? `${base}..${localSha}` : `${emptyTreeOid()}..${localSha}`;
} else {
// Existing branch (incl. force-push): net new content remote..local.
range = `${remoteSha}..${localSha}`;
}
// -U0: only changed lines; we keep lines starting with '+' (added), drop the
// +++ file header. Unified diff added lines start with a single '+'.
// Strict (#1946): a failed diff used to return "" and the push sailed
// through unscanned — fail open on the exact path the guard exists for.
//
// --no-ext-diff: a user's `diff.external` driver replaces the entire diff
// with its own output — with one set, `git diff` emits zero '+' lines, so an
// unhardened scanner reads an empty diff and exits 0 on a push full of
// secrets. Reachable from ordinary user config, not hypothetical. (#2498)
// --no-textconv: a .gitattributes textconv driver can likewise rewrite
// content before we ever see it. (#2498)
const diff = gitStrict([
"diff", "--unified=0", "--no-color", "--no-ext-diff", "--no-textconv",
"--src-prefix=a/", "--dst-prefix=b/", range,
]);
return collectAddedLines(diff);
}
/**
* Added-line text from a unified diff. Shared by both range strategies so the
* hunk-aware header handling below cannot drift between them.
*/
function collectAddedLines(diff: string, commit?: string): AddedFile[] {
const added: AddedFile[] = [];
// Hunk-aware header skip (#2498): `+++ ` is only a FILE HEADER outside a
// hunk. Inside a hunk, an added content line whose text begins with "++"
// renders as "+++<content>" — the old blanket startsWith("+++") skip
// silently dropped exactly those lines from the scan.
let inHunk = false;
let file = "";
let fileLines: string[] = [];
let lineNumbers: number[] = [];
// File line tracking from hunk headers. A combined (--cc) header has one
// more `@` than it has parents, and each body line one prefix column per
// parent; a line belongs to the resulting file unless a column says `-`.
let parents = 1;
let nextLine = 0;
const flush = () => {
const lines = releaseVersionExempt(file, fileLines);
if (lines.length) added.push({ file, text: lines.join("\n"), lines: lineNumbers, ...(commit ? { commit } : {}) });
fileLines = [];
lineNumbers = [];
};
for (const line of diff.split("\n")) {
// `diff --` rather than `diff --git`: a merge scanned with --cc emits
// `diff --cc <path>`, so a --git-only reset left inHunk true across file
// boundaries and read the next file's `+++ b/...` header as content. Only
// noise (it over-scans, never under-scans), but the boundary is real.
if (line.startsWith("diff --")) { flush(); file = ""; inHunk = false; continue; }
if (line.startsWith("@@")) {
inHunk = true;
parents = (/^@+/.exec(line)?.[0].length ?? 2) - 1;
nextLine = Number(/ \+(\d+)(?:,\d+)? @+/.exec(line)?.[1] ?? 0);
continue;
}
if (!inHunk && line.startsWith("+++ ")) { file = line.startsWith("+++ b/") ? line.slice(6) : ""; continue; }
if (!inHunk && (line.startsWith("+++") || line.startsWith("---"))) continue;
const inResult = inHunk && !line.startsWith("\\") && !line.slice(0, parents).includes("-");
const lineNumber = inResult ? nextLine++ : 0;
if (line.startsWith("+")) {
// A combined (--cc) body line carries one prefix column per parent.
fileLines.push(line.slice(inHunk ? parents : 1));
lineNumbers.push(lineNumber);
}
}
flush();
return added;
}
const VERSION_FILE_CONTENT = /^\d+(?:\.\d+)+$/;
const CHANGELOG_RELEASE_VERSION = /^(#{1,6}[\t ]*\[?v?)\d+(?:\.\d+){3}(?=[\]\t ]|$)/;
/**
* #2856: a 4-part release version (MAJOR.MINOR.PATCH.MICRO) is shaped exactly like a
* public IPv4 address, so /ship's own VERSION bump and CHANGELOG heading
* raised pii.ip_public on every release. Exempt only those two release
* artifacts, decided from the pushed diff alone (works in shallow clones):
* a root VERSION whose entire added content is one version string, and the
* version token of a root CHANGELOG.md release heading (the rest of the
* heading is still scanned). Every other file — and any other position in
* these two — is scanned unchanged; the engine's global allowlist is untouched.
*/
function releaseVersionExempt(file: string, lines: string[]): string[] {
if (file === "VERSION" && lines.length === 1 && VERSION_FILE_CONTENT.test(lines[0].trim())) return [];
if (file === "CHANGELOG.md") return lines.map((l) => l.replace(CHANGELOG_RELEASE_VERSION, "$1VERSION"));
return lines;
}
/**
* Byte budget per scan() call. Kept comfortably under redact-engine's
* DEFAULT_MAX_BYTES (1 MiB) so a slice never trips its oversize guard.
*/
const SCAN_CHUNK_BYTES = 768 * 1024;
const SCAN_OVERLAP_CHARS = 16 * 1024;
const MAX_SCAN_BYTES = 1024 * 1024;
/**
* Scan complete-line core slices with bounded left and right context. Admit
* only findings whose original captured start lies inside that core.
*
* Why: the engine refuses input over its byte cap and fails closed, which is
* right for one scan() call but wrong as a push policy — a feature branch
* catching up to a busy main legitimately produces more added lines than the
* cap (1,146,782 bytes against the 1 MiB default in the push that prompted
* this, and only ~7% of that was the lockfile). The push then blocked on
* `engine.input_too_large` — a size error naming no credential — which trains
* people to reach for --no-verify, defeating the guardrail far more thoroughly
* than a large diff does.
*
* A match may need context on either adjacent line (`nearWindow` reaches 300
* normalized characters). The full core lines keep opaque spans intact; the
* engine's normalization map selects context in normalized units without
* changing the raw input. A partial context may create an artificial boundary,
* but its findings are never admitted as core findings.
*
* Fail-closed is preserved: a SINGLE line over the budget is still passed to
* the engine intact, so a genuinely unscannable blob (minified bundle,
* embedded base64) trips input_too_large and blocks exactly as before.
*
* The engine reports line/col relative to each slice; an admitted finding's
* `line` is rewritten to its 1-based line within `added`, which the caller
* maps to the file line. Do not lift this into the engine, where callers rely
* on absolute line numbers.
*/
function scanAddedLines(added: string, opts: Parameters<typeof scan>[1]): Finding[] {
const findings: Finding[] = [];
const cores: Array<{ start: number; end: number }> = [];
const lines = added.split("\n");
let start = 0;
let end = 0;
let bytes = 0;
for (let i = 0; i < lines.length; i++) {
const line = lines[i];
const newline = i < lines.length - 1 ? 1 : 0;
const lineBytes = Buffer.byteLength(line, "utf8") + newline;
if (bytes && bytes + lineBytes > SCAN_CHUNK_BYTES) {
cores.push({ start, end });
start = end;
bytes = 0;
}
end += line.length + newline;
bytes += lineBytes;
if (lineBytes > SCAN_CHUNK_BYTES) {
cores.push({ start, end });
start = end;
bytes = 0;
}
}
if (bytes) cores.push({ start, end });
const contextStart = (boundary: number): number => {
if (!boundary) return 0;
let length = Math.min(boundary, SCAN_OVERLAP_CHARS * 2);
while (true) {
const from = boundary - length;
const raw = added.slice(from, boundary);
if (Buffer.byteLength(raw, "utf8") > MAX_SCAN_BYTES) return from;
const { normalized, map } = normalizeWithMap(raw);
if (!from || normalized.length > SCAN_OVERLAP_CHARS) {
return from + (map[Math.max(0, normalized.length - SCAN_OVERLAP_CHARS - 1)] ?? 0);
}
length = Math.min(boundary, length * 2);
}
};
const contextEnd = (boundary: number): number => {
if (boundary === added.length) return boundary;
let length = Math.min(added.length - boundary, SCAN_OVERLAP_CHARS * 2);
while (true) {
const to = boundary + length;
const raw = added.slice(boundary, to);
if (Buffer.byteLength(raw, "utf8") > MAX_SCAN_BYTES) return to;
const { normalized, map } = normalizeWithMap(raw);
if (to === added.length || normalized.length > SCAN_OVERLAP_CHARS) {
return boundary + (map[Math.min(normalized.length, SCAN_OVERLAP_CHARS + 1)] ?? length);
}
length = Math.min(added.length - boundary, length * 2);
}
};
const addedLineStarts = [0];
for (let i = 0; i < added.length; i++) if (added[i] === "\n") addedLineStarts.push(i + 1);
const lineAt = (offset: number): number => {
let lo = 0;
let hi = addedLineStarts.length - 1;
while (lo < hi) {
const mid = (lo + hi + 1) >> 1;
if (addedLineStarts[mid] <= offset) lo = mid;
else hi = mid - 1;
}
return lo + 1;
};
for (const core of cores) {
const left = contextStart(core.start);
const text = added.slice(left, contextEnd(core.end));
const result = scan(text, opts);
if (result.oversize) {
findings.push(...result.findings);
return findings;
}
const starts = [0];
for (let i = 0; i < text.length; i++) if (text[i] === "\n") starts.push(i + 1);
for (const finding of result.findings) {
const offset = starts[finding.line - 1] + finding.col - 1 + left;
if (offset >= core.start && offset < core.end) findings.push({ ...finding, line: lineAt(offset) });
}
}
return findings;
}
function logSkip(reason: string): void {
try {
const home = resolveStateRoot();
const dir = path.join(home, "security");
// mkdirpSync, not bare mkdirSync: bun-on-Windows EEXIST (#2635). This site
// is try-wrapped by the caller, so the old failure was a silent skip-log
// loss rather than a crash — the fix makes the log survive, not un-crash.
mkdirpSync(dir);
fs.appendFileSync(
path.join(dir, "prepush-skip.jsonl"),
JSON.stringify({ ts: new Date().toISOString(), reason }) + "\n",
);
} catch {
// best-effort; never block a push because logging failed
}
}
const HISTORY_MAX_COMMITS = 50_000;
const HISTORY_TIMEOUT_MS = 5_000;
function configValues(args: string[]): string[] {
return git(["config", ...args]).split("\n").map((v) => v.trim()).filter(Boolean);
}
/**
* Addresses `pii.email` treats as not a new leak (#3060): the pusher's own
* (`git config user.email`), the ones listed in `gstack.redact.allowEmail`,
* and the author and committer addresses (raw and mailmapped) the destination
* will hold after this push. That last set is the history of every pushed tip
* (pushed commits become public with the push, third-party authors included),
* of each remote tip being replaced, and, for a ref whose remote tip is new or
* unknown, the push remote's own tracking refs. A URL or unknown destination
* never borrows a remote's tracking refs; another remote's history never counts.
*
* Each input fails on its own and a failure omits only that input, which is
* the behavior before these exemptions existed. The history read is bounded:
* past the commit cap it uses the newest authors it read; an error or the
* timeout passes no history, and the reason is returned so the report can say
* suppression was limited.
*/
function emailScanOptions(historyRevs: string[], trackingRemote: string | null): {
options: Pick<ScanOptions, "selfEmail" | "repoPublicEmails">;
limited?: string;
} {
const selfEmail = configValues(["--get", "user.email"])[0];
const allowed = configValues(["--get-all", "gstack.redact.allowEmail"]);
const r = spawnSync("git", [
"log", "--no-show-signature", "--no-color", `--max-count=${HISTORY_MAX_COMMITS}`, "--format=%ae%n%aE%n%ce%n%cE",
...historyRevs, ...(trackingRemote ? [`--remotes=${trackingRemote}/*`] : []), "--",
], { encoding: "utf8", maxBuffer: 64 * 1024 * 1024, timeout: HISTORY_TIMEOUT_MS, killSignal: "SIGKILL" });
const timedOut = (r.error as NodeJS.ErrnoException | undefined)?.code === "ETIMEDOUT";
const history = r.status === 0 && !r.error
? (r.stdout ?? "").split("\n").map((e) => e.trim()).filter((e) => e.includes("@"))
: null;
return {
options: {
...(selfEmail ? { selfEmail } : {}),
repoPublicEmails: [...new Set([...(history ?? []), ...allowed])],
},
...(history ? {} : { limited: timedOut ? `reading commit history took over ${HISTORY_TIMEOUT_MS / 1000} s` : "reading commit history failed" }),
};
}
const ALLOWLIST_PATH = ".gstack-redact-allowlist";
const ALLOWLIST_MAX_BYTES = 64 * 1024;
/**
* Entries of the allowlist committed in `sha`. Read from the pushed commit so
* a suppression exists only as a reviewable commit. Absent, unreadable or
* oversized means no entries: the file can record a reviewed exception, never
* loosen the scan by accident.
*/
function committedAllowlist(sha: string): string[] {
const spec = `${sha}:${ALLOWLIST_PATH}`;
const size = Number(git(["cat-file", "-s", spec]).trim());
if (!size) return [];
if (size > ALLOWLIST_MAX_BYTES) {
process.stderr.write(
`gstack-redact-prepush: ${ALLOWLIST_PATH} in ${sha.slice(0, 12)} is over ${ALLOWLIST_MAX_BYTES} bytes; ignoring it.\n`,
);
return [];
}
return git(["cat-file", "blob", spec]).split("\n")
.map((l) => l.trim())
.filter((l) => l && !l.startsWith("#"));
}
/** File line (and commit) of a finding. Never the matched value. */
function findingWhere(f: Finding, part: AddedFile): string {
const line = part.lines[f.line - 1];
return `${part.file || "(unknown file)"}${line ? `:${line}` : ""}` +
(part.commit ? ` in commit ${part.commit.slice(0, 12)}` : "");
}
/** One MEDIUM report line: rule id, file line (and commit), and the fix. Never the matched value. */
function mediumLine(f: Finding, part: AddedFile): string {
const where = findingWhere(f, part);
const fix = f.id === "pii.email"
? "remove the address, or allow it: git config --add gstack.redact.allowEmail <address>"
: "remove the value before it becomes public";
return ` MEDIUM ${f.id} ${where} fix: ${fix}\n`;
}
function main() {
if ((process.env.GSTACK_REDACT_PREPUSH || "").toLowerCase() === "skip") {
logSkip(process.env.GSTACK_REDACT_PREPUSH_REASON || "env-skip");
process.stderr.write("gstack-redact-prepush: skipped via GSTACK_REDACT_PREPUSH=skip\n");
process.exit(0);
}
const stdin = fs.readFileSync(0, "utf8");
const refs = stdin
.split("\n")
.map((l) => l.trim())
.filter(Boolean)
.map((l) => l.split(/\s+/));
const parts: Array<{ part: AddedFile; allowlist: string[] }> = [];
const allowEntries = new Set<string>();
const historyRevs = new Set<string>();
let trackingRemote: string | null = null;
for (const fields of refs) {
// Fail CLOSED on a ref line we cannot parse (#2498): git hands pre-push
// exactly "<local ref> <local sha> <remote ref> <remote sha>" — anything
// else means we cannot tell WHAT is being pushed, and silently skipping
// it would leave that ref unscanned.
const [, localSha, , remoteSha] = fields;
const shaShaped = (s: string | undefined) => !!s && /^[0-9a-f]{40,64}$/i.test(s);
if (fields.length !== 4 || !shaShaped(localSha) || !shaShaped(remoteSha)) {
process.stderr.write(
"\n⛔ gstack-redact-prepush BLOCKED the push — could not parse a pre-push ref line, " +
"so its content cannot be scanned.\n" +
` line: ${JSON.stringify(fields.join(" "))}\n` +
"Bypass if you're sure: GSTACK_REDACT_PREPUSH=skip git push (or git push --no-verify)\n",
);
process.exit(1);
}
if (ZERO.test(localSha!)) continue; // branch delete → nothing pushed
let added: AddedFile[];
try {
added = addedLinesFor(localSha, remoteSha || "0");
} catch (err) {
// Fail CLOSED (#1946): if we can't compute the pushed diff we can't
// scan it, and unscanned-but-allowed is the failure mode this hook
// exists to prevent.
process.stderr.write(
"\n⛔ gstack-redact-prepush BLOCKED the push — could not compute the pushed diff, " +
"so it cannot be scanned for credentials.\n" +
` (${err instanceof Error ? err.message.split("\n")[0] : String(err)})\n` +
"Bypass if you're sure: GSTACK_REDACT_PREPUSH=skip git push (or git push --no-verify)\n",
);
process.exit(1);
}
const allowlist = committedAllowlist(localSha);
for (const entry of allowlist) allowEntries.add(entry);
for (const part of added) if (part.text.trim()) parts.push({ part, allowlist });
historyRevs.add(localSha);
if (!ZERO.test(remoteSha) && objectExists(remoteSha)) historyRevs.add(remoteSha);
else {
const target = pushTarget();
if (target.kind === "remote") trackingRemote = target.name;
}
}
const allHigh: Finding[] = [];
const mediums: string[] = [];
let emailMedium = false;
const emails = parts.length ? emailScanOptions([...historyRevs], trackingRemote) : { options: {} };
// Visibility doesn't change HIGH behavior; pass private so nothing is treated
// as public-strict (HIGH blocks regardless either way).
// Sliced (see scanAddedLines) so a large-but-legitimate diff is actually
// scanned rather than blocked unscanned on the engine's size cap.
// Scanned per file so a pattern can use the file's type (#3048).
// An allowlisted span is found by scanning with and without the entries,
// so the report names exactly what they suppressed.
const allowed: string[] = [];
const findingKey = (f: Finding) => `${f.id}:${f.line}:${f.col}`;
for (const { part, allowlist } of parts) {
const opts = { repoVisibility: "private" as const, ...emails.options, ...(part.file ? { sourcePath: part.file } : {}) };
const findings = scanAddedLines(part.text, opts);
const kept = allowlist.length
? new Set(scanAddedLines(part.text, { ...opts, allowlist }).map(findingKey))
: null;
for (const f of findings) {
if (kept && !kept.has(findingKey(f))) {
if (f.severity === "HIGH" || f.severity === "MEDIUM") allowed.push(` ${f.severity} ${f.id} ${findingWhere(f, part)}\n`);
continue;
}
if (f.severity === "HIGH") allHigh.push(f);
else if (f.severity === "MEDIUM") {
mediums.push(mediumLine(f, part));
if (f.id === "pii.email") emailMedium = true;
}
}
}
if (allowEntries.size > 0) {
process.stderr.write(
`gstack-redact-prepush: ${ALLOWLIST_PATH} (${allowEntries.size} entr${allowEntries.size === 1 ? "y" : "ies"}) ` +
`suppressed ${allowed.length} finding(s) in this push${allowed.length ? ":" : "."}\n`,
);
for (const line of allowed) process.stderr.write(line);
}
if (mediums.length > 0) {
process.stderr.write(
`gstack-redact-prepush: ${mediums.length} MEDIUM finding(s) in pushed diff (PII/internal). ` +
"Not blocking. Review before this becomes public.\n",
);
for (const line of mediums) process.stderr.write(line);
if (emailMedium && "limited" in emails && emails.limited) {
process.stderr.write(
`gstack-redact-prepush: existing-email suppression was limited for this push (${emails.limited}), ` +
"so an address already in the history may be listed above.\n",
);
}
}
if (allHigh.length > 0) {
// A scan that could not RUN is not a scan that FOUND something. Reporting
// "credential(s) in the pushed diff — rotate the credential" for an
// `engine.*` finding tells the operator to rotate a secret that was never
// detected, on a diff that was never read. Blocking is still right (fail
// closed), but the reason must be the true one: a guardrail that cries wolf
// is a guardrail that gets bypassed by reflex, which is worse than none.
// The computed range can be much larger than the visible tip diff, and a
// single over-cap line is deliberately withheld rather than partly scanned.
const unscanned = allHigh.filter((f) => f.id.startsWith("engine."));
const secrets = allHigh.filter((f) => !f.id.startsWith("engine."));
if (secrets.length > 0) {
process.stderr.write(
"\n⛔ gstack-redact-prepush BLOCKED the push — credential(s) in the pushed diff:\n\n",
);
for (const f of secrets) {
process.stderr.write(` HIGH ${f.id} ${f.preview}\n`);
}
process.stderr.write(
"\nRotate the credential (a pushed secret is compromised) and remove it from the diff.\n" +
`If the value is a known-benign sample, commit its exact matched text as a line of ${ALLOWLIST_PATH}\n` +
"at the repo root instead of reshaping code to dodge the pattern; every push lists what that file suppresses.\n",
);
}
if (unscanned.length > 0) {
process.stderr.write(
"\n⛔ gstack-redact-prepush BLOCKED the push — the diff could NOT be scanned.\n" +
" No credential was found; none was looked for. Blocking fail-closed.\n\n",
);
for (const f of unscanned) {
process.stderr.write(` ${f.id}: ${f.description}\n`);
}
process.stderr.write(
"\nA long line, large first push, or unavailable remote tip may exceed the\n" +
"per-slice safety cap. Check the destination and range; scan the diff yourself\n" +
"before bypassing: `git diff <base>..HEAD | grep -inE \'password|secret|token|api.?key\'`.\n",
);
}
process.stderr.write(
"This is a guardrail: `git push --no-verify` or `GSTACK_REDACT_PREPUSH=skip git push` bypass it.\n",
);
process.exit(1);
}
process.exit(0);
}
main();