Files
gstack/test/pr-title-sync-workflow-safety.test.ts
T
f30b7b788a v1.91.13.0 fix: severe bug wave (autoplan guard, land-and-deploy CI gate, project identity, gbrain, review gates) (#3013)
* fix(land-and-deploy): never merge over red, pending or missing CI (#2995)

/land-and-deploy gated only on required checks, so a repo with none
(every GitHub Free private repo, any repo without branch protection)
skipped the CI wait and could merge over red CI. The verdict now comes
from bin/gstack-ci-gate, a tested helper bound to the head SHA: all
checks by default, cancel/unknown buckets fail, NO_CHECKS needs explicit
per-head approval, and both merge commands (including --auto) run only
after a PASS for the same SHA they pin with --match-head-commit.

Co-authored-by: harshitgavita-07 <harshit.gavita@gmail.com>

* fix(slug): give nested-group projects their own state bucket and verify checkpoint identity (#3003, #2901)

gstack-slug kept only the last two remote path segments, so
customer-a/product/repo and customer-b/product/repo shared one bucket
and /context-restore could present another project's checkpoint.

- One canonical-remote rule (bash + TS, golden vectors) used by all six
  resolvers: lowercased host, no userinfo/port/.git, Azure DevOps and
  Bitbucket Server normalized. 3+ segment remotes get
  <last-two>-<16 hex>; two-segment remotes (all of GitHub) are unchanged.
- Versioned sticky cache (v2:) keeps cache hits git-free.
- context-save stamps the credential-free remote and root;
  context-restore only treats this project's checkpoints as latest,
  reports foreign/unverified ones first, and treats a different root
  with the same remote (Conductor handoff) as information.
- gstack-slug --adopt-legacy copies only files the user confirms.
- question-preference-hook reads the same versioned cache, so never-ask
  preferences fire; unknown identity asks instead of auto-deciding.

* fix(gbrain): brain context loads, managed sync works, memory sync stops duplicating (#2883, #2985, #2922)

- brain-context-load called 'gbrain list_pages' (an MCP tool name), so
  every skill's brain context was 0 bytes. It now uses 'gbrain list'
  with gbrain's real filter flags, and prints one line when queries fail.
- The code-stage sync always passes --no-pull (managed gbrain >= 0.51
  refuses without it); a gbrain that rejects the flag fails with an
  upgrade message and is never retried without it. stderr is streamed
  with a bounded tail and quoted in failure summaries.
- /sync-gbrain gains --sources (and GSTACK_MEMORY_INGEST_SOURCES); by
  default it skips types already synced through a federated curated
  source, so curated pages stop being re-duplicated.

Co-authored-by: Vitor Cepeda Lopes <vitor@vitorcepedalopes.com>
Co-authored-by: harshitgavita-07 <harshit.gavita@gmail.com>

* fix(codex,cso): clean Codex reviews stop failing; semgrep and trivy SARIF results import (#2769, #3011)

- Codex review gate is PASS / FAIL / UNVERIFIED. Untagged output (what
  a clean 'codex review --base' prints) is UNVERIFIED, not FAIL, and
  never a false PASS; P2/P3-only output passes.
- import-sarif resolves semgrep's undeclared %SRCROOT% and trivy's
  ROOTPATH (including git-URL targets) through explicit, captured
  mappings; declared bases must sit inside the source root or the
  run's checkout, URIs are percent-decoded, and rejected results are
  counted with rule, base and path instead of silently dropped.

* fix: version strings are not IPs, failed update checks are not 'up to date', private security reporting (#2856, #2786, #2984)

- The pre-push redact hook exempts a root VERSION file that is only a
  version string and the version token in CHANGELOG release headings;
  the engine allowlist is untouched.
- gstack-update-check caches an unreadable remote as CHECK_FAILED with
  a short TTL instead of UP_TO_DATE; /gstack-upgrade never reports
  'latest' after a failed check.
- SECURITY.md points reporters at GitHub private vulnerability
  reporting.
- pr-title-sync (pull_request_target, write token) pins actions/checkout
  to a reviewed SHA.

* fix(autoplan): publication guard accepts real Claude Code journals (#3007, #3009, #2968, #2977, #2986, #3006)

The guard required a session journal to start with a user message, but
Claude Code starts it with SessionStart hook attachments (and a forked
compacted session starts with a compact_boundary), so every phase was
denied for anyone with a SessionStart hook. Windows path spellings
(C:/, /c/, drive case) were rejected too.

- Reader accepts the captured startup, compact, clear and fork shapes
  (claude 2.1.284 fixture), computes root uniqueness over every
  candidate, and returns typed reason codes instead of one string.
- Hard denials (competing root, foreign cwd, sidechain, agent, cycle,
  malformed, identity) say what happened and how to run the reviews by
  hand; unflushed journals still retry.
- A well-formed, stable journal whose root shape is unrecognized now
  warns ('phase publication was NOT verified') and continues, without
  overriding Claude Code's own permission check; no bypass switch.
- Windows paths fold before validation, so '..' is still rejected.
- Periodic drift canary runs the reader on fresh journals from the
  pinned claude.

Co-authored-by: Charles M Berg <77517285+CharlesBerg85@users.noreply.github.com>
Co-authored-by: laurenbdaniels-coder <264343335+laurenbdaniels-coder@users.noreply.github.com>
Co-authored-by: Jordan Mayer <176445995+maydaycyber@users.noreply.github.com>
Co-authored-by: elmesery-dev <elmesery.dev@gmail.com>

* v1.91.13.0: severe bug fix wave release notes, stale gbrain verbs in docs

* test: seed the versioned slug cache in the memory-injection deny case (#2901)

The preference hook now resolves the project bucket from gstack-slug's
v2 slug-cache entry instead of basename(cwd); this fixture still wrote
preferences under the basename bucket without the entry the skill
preamble always writes first.

* fix: full-suite follow-ups for the fix wave

- gstack-slug's --help heredoc is over 512B, so it gets the BASH_COMPAT=50
  pipe-deadlock guard like the other bins.
- Refresh the codex/factory ship goldens for the plan-gates slug change.
- The periodic lane gains the autoplan journal drift canary (107 paid
  files); raise the eval-slices cap to 25 so every slice still starts
  at once until its recorded duration lets the planner pack it.

* fix(plan-eng-review): clearer Step 0 sequencing, write routes and section outputs

The workflow judge scored plan-eng-review's instructions at
actionability 3.67 (bar 4) on text unchanged from main. Resolve the
ambiguities it named together, without changing any gate or output:
Step 0 names the Read's order and keeps one STOP; the pending-call,
read-only, forbidden-write and failed-save rules become one Write
routes block; header/label limits and grid checks are concrete;
Performance gets a checklist; each section has a stated output shape;
Lake Score is defined. Same total length (parity cap).

---------

Co-authored-by: harshitgavita-07 <harshit.gavita@gmail.com>
Co-authored-by: Vitor Cepeda Lopes <vitor@vitorcepedalopes.com>
Co-authored-by: Charles M Berg <77517285+CharlesBerg85@users.noreply.github.com>
Co-authored-by: laurenbdaniels-coder <264343335+laurenbdaniels-coder@users.noreply.github.com>
Co-authored-by: Jordan Mayer <176445995+maydaycyber@users.noreply.github.com>
Co-authored-by: elmesery-dev <elmesery.dev@gmail.com>
2026-10-02 11:11:00 -07:00

143 lines
5.7 KiB
TypeScript

/**
* pr-title-sync.yml is a `pull_request_target` workflow — static injection
* tripwire (gate, free).
*
* The anxiety this kills: `pull_request_target` runs with a WRITE token in the
* base-repo context, even for fork PRs. That is what lets this workflow rewrite
* fork-PR titles (the backstop). It is also the single most dangerous workflow
* trigger in GitHub Actions. Two classic footguns turn it into remote code
* execution / token theft, and `actionlint` catches NEITHER:
*
* 1. Checking out the PR head (`actions/checkout` with a `ref:` pointing at
* `pull_request.head` / `head_ref`) and then running anything from it —
* that executes attacker-controlled fork code with the write token.
* 2. Interpolating an attacker-controlled `${{ github.event.pull_request.* }}`
* field directly INSIDE a `run:` block — the title/body are attacker-
* controlled and the `${{ }}` is expanded into the shell before execution,
* so a crafted title runs as code. Those fields MUST arrive via `env:` and
* be referenced as `"$VAR"` (shell-quoted), never inlined.
*
* This tripwire reads the workflow file directly and fails CI if either pattern
* reappears. Mirrors the static-grep invariant tests in browse/test
* (terminal-agent-pid-identity, server-sanitize-surrogates).
*
* Note: `gh api ... -q '.head.sha'` inside a run block is SAFE (reading PR
* metadata as data via a jq filter string, not `${{ }}` interpolation), so we
* ban the interpolation form specifically, not the literal substring `head.sha`.
*/
import { describe, test, expect } from 'bun:test';
import * as fs from 'node:fs';
import * as path from 'node:path';
const WORKFLOW = path.resolve(__dirname, '..', '.github', 'workflows', 'pr-title-sync.yml');
/** Indentation width (count of leading spaces) of a line. */
function indent(line: string): number {
const m = line.match(/^( *)/);
return m ? m[1].length : 0;
}
/**
* Return the lines that live inside a `run:` block, each tagged with its 1-based
* line number. Handles both `run: |` (multiline) and `run: <inline command>`.
*/
function runBlockLines(content: string): Array<{ n: number; text: string }> {
const lines = content.split('\n');
const out: Array<{ n: number; text: string }> = [];
let inRun = false;
let runIndent = -1;
for (let i = 0; i < lines.length; i++) {
const line = lines[i];
const n = i + 1;
const inlineRun = line.match(/^(\s*)run:\s*(\S.*)$/); // `run: echo foo`
const blockRun = /^(\s*)run:\s*(\|>?[+-]?)?\s*$/.test(line); // `run: |`
if (inlineRun && !/^\|/.test(inlineRun[2])) {
out.push({ n, text: inlineRun[2] });
inRun = false;
continue;
}
if (blockRun) {
inRun = true;
runIndent = indent(line);
continue;
}
if (inRun) {
if (line.trim() === '') {
out.push({ n, text: line });
continue;
}
// Block ends when a non-empty line is indented at or below the `run:` key.
if (indent(line) <= runIndent) {
inRun = false;
} else {
out.push({ n, text: line });
}
}
}
return out;
}
describe('pr-title-sync.yml pull_request_target safety', () => {
const content = fs.readFileSync(WORKFLOW, 'utf-8');
test('workflow file exists', () => {
expect(fs.existsSync(WORKFLOW)).toBe(true);
});
test('does NOT check out the PR head ref (no fork-code execution)', () => {
const offenders: string[] = [];
content.split('\n').forEach((line, i) => {
// A checkout `ref:` (or any `ref:`) pointing at the PR head is the footgun.
if (/ref:\s*\$\{\{[^}]*(pull_request\.head|head_ref)/.test(line)) {
offenders.push(` L${i + 1}: ${line.trim()}`);
}
});
if (offenders.length > 0) {
throw new Error(
`pr-title-sync.yml checks out the PR head under pull_request_target — that ` +
`runs attacker-controlled fork code with a write token. Check out the base ` +
`repo (no ref:) and read PR-head data via the API instead.\n` +
offenders.join('\n'),
);
}
});
test('does NOT interpolate ${{ github.event.pull_request.* }} inside a run: block', () => {
const offenders: string[] = [];
for (const { n, text } of runBlockLines(content)) {
if (/\$\{\{\s*github\.event\.pull_request/.test(text)) {
offenders.push(` L${n}: ${text.trim()}`);
}
}
if (offenders.length > 0) {
throw new Error(
`pr-title-sync.yml inlines an attacker-controlled PR field into a run: block ` +
`— a crafted PR title/body executes as shell. Pass it via env: and ` +
`reference "$VAR" (shell-quoted) instead.\n` +
offenders.join('\n'),
);
}
});
test('uses pull_request_target (the hardening is actually present)', () => {
// Positive assertion: if someone reverts to plain pull_request, the fork
// backstop silently stops working (read-only token). Keep it intentional.
expect(/^on:\s*$/m.test(content) || /\bpull_request_target\b/.test(content)).toBe(true);
expect(content).toMatch(/\bpull_request_target\b/);
});
test('passes the PR title through env:, not raw interpolation', () => {
// The safe pattern: OLD_TITLE: ${{ github.event.pull_request.title }} in an
// env: mapping, consumed as "$OLD_TITLE" in script.
expect(content).toMatch(/env:/);
expect(content).toMatch(/github\.event\.pull_request\.title/);
});
test('every action runs at a reviewed commit SHA, not a movable tag', () => {
// The job holds a write token; a retagged action would run with it.
const uses = [...content.matchAll(/^\s*(?:-\s*)?uses:\s*(\S+)/gm)].map((m) => m[1]);
expect(uses.length).toBeGreaterThan(0);
for (const ref of uses) expect(ref).toMatch(/^[\w.-]+\/[\w.-]+@[0-9a-f]{40}$/);
});
});