[dirtysouthalpha]

Summary

• README: H1 + bold one-line pitch under the banner, CI badge for .github/workflows/ci.yml, Quick start → Quickstart, new ## Tested on table built only from INSTALL.md / package.json engines / the CI workflow (no invented numbers). Version/test badges and the stale "347 tests" comment refreshed to 3.3.0 / 866. All existing content kept; real screenshots in assets/ retained.
• Tests: existing vitest suite was already green (855). Added two tests for previously untested pure modules: tests/compression.test.ts (local fallback compression, cache, stats) and tests/project-detector.test.ts (node/rust/python/go detection, name fallback, priority, icon). No source changes were needed.
• CI: rewritten to a single job on ubuntu-24.04 (no matrix), actions/setup-node@v4 Node 22 with cache: npm covering both lock files, npm ci → lint → test → build → GUI npm ci && npm run build. paths-ignore, concurrency cancel-in-progress, permissions: contents: read, PUPPETEER_SKIP_DOWNLOAD=1 (CI never drives a browser).

Test run

PUPPETEER_SKIP_DOWNLOAD=1 npm ci && npm test
 Test Files  108 passed (108)
      Tests  866 passed (866)

Also run locally: npm run lint (clean), npm run build (ok), cd gui && npm ci && npm run build (ok, ~1s). Tauri build not run locally. Note: plain npm ci fails on this machine because Puppeteer's Chromium download is blocked by the sandbox — hence the env var, which CI now sets too.

Scan findings

Scan findings (working tree + full git history; nothing auto-fixed, history not rewritten)

• No real credentials found. Test fixtures in tests/ (sync, secrets-*, redaction) use obviously fake tokens (e.g. a 20-char "sk-ant-..." string, "AKIA..." examples) — fine, but they will trip naive secret scanners; consider adding a .gitleaks.toml / allowlist later.
• IPs in tests (10.0.0.1, 172.16.0.1, 192.168.1.5) are SSRF-guard test inputs, not infrastructure.
• No fleet hostnames, no Premier-related content found.

Git history was not rewritten and nothing was auto-fixed; this PR only adds/edits README.md, two test files, and ci.yml.

Notes / follow-ups

• README.md still says "New in 3.2" while package.json is 3.3.0 — left as-is (content decision for the maintainer).
• gui/package.json is at 3.2.0 vs root 3.3.0; not touched.
• Consider a .gitleaks.toml allowlist for the fake tokens in test fixtures.

🤖 Generated with Claude Code

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-11-01T13:56:40.978091Z	dd41a40	PR opened

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: dd41a406fd

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Keep CI enabled for functional Markdown changes

When a push or PR changes only a built-in file such as src/commands/builtin/ship.md, **.md matches nested Markdown files too, so the workflow skips CI. These files define runtime commands and are copied into dist by the build, so changes to commands, agents, or skills receive no test or build check. Limit the exclusion to documentation paths; GitHub’s path-pattern docs confirm that **.md matches Markdown files throughout the repository.

Useful? React with 👍 / 👎.