AGENTS

Canonical skill index for agent-compatible tooling in this repository.

Quality tooling

Prefer uv run lintro chk or uv run lintro fmt when lintro is available; only use bundled native tools (for example bandit, ruff, black) when lintro is unavailable or a skill here explicitly overrides that preference.

Skills

• analyze-code: Code-level quality analysis. Use when asked to review code for smells, security issues, implementation quality, or test coverage. (skills/analyze-code/SKILL.md)
• analyze-project: High-level project analysis. Use when asked to analyze, review, or evaluate a project's architecture, structure, and overall health. (skills/analyze-project/SKILL.md)
• analyze-tests: Test suite analysis. Use when asked to analyze, review, or evaluate a project's tests for quality, coverage gaps, and best practices. (skills/analyze-tests/SKILL.md)
• audit-merges: Deep retrospective audit of everything merged to main over a window - code quality against the stand-* skills, introduced bugs, unsafe changes, merges over red signals, hollow thread resolutions, and untracked deferrals - with a tabbed HTML report as the deliverable. Use when asked to audit merges, review what landed on main, or verify automated/agent-authored merges were up to standard. (skills/audit-merges/SKILL.md)
• babysit-pr: Autonomously drive an open PR to merge-ready state by triaging Greptile and CodeRabbit review comments, fixing CI failures, handling CodeRabbit rate limits, and looping until checks are green with no unresolved actionable threads. With the --merge flag, also merge the PR(s) once binding merge-queue conditions are met. Use when asked to babysit a PR, shepherd a PR, or keep a PR merge-ready until review/CI cycles complete. (skills/babysit-pr/SKILL.md)
• backlog: Interactive dispatcher for backlog work. Asks one routing question — drive PRs to merge, or implement issues to open-PR state — then follow-ups based on the answer, and hands off to implement-issues / babysit-pr with standing policy applied. Use when asked to work the backlog or run /backlog. (skills/backlog/SKILL.md)
• branch: Start work on a new branch or worktree. Use when asked to start a new branch, new worktree, begin work on a feature/fix, or start fresh. Supports issue numbers and plain descriptions. (skills/branch/SKILL.md)
• coderabbit: Run CodeRabbit CLI for pre-push AI diff review. Use when asked for CodeRabbit, cr review, or as part of the default pre-push review set with greptile and lintro-review. Max 2-3 runs per change set. (skills/coderabbit/SKILL.md)
• commit: Pre-commit workflow and commit guidelines. Use when asked to commit changes. Requires passing lint and tests, signed commits, semantic prefixes, imperative mood. (skills/commit/SKILL.md)
• greptile: Run Greptile CLI for pre-push AI branch review. Use when asked for Greptile review, raycast/extensions pre-PR checks, or as part of the default pre-push review set with coderabbit and lintro-review. Max 2 runs per change set. (skills/greptile/SKILL.md)
• implement-issues: Implement a set of GitHub issues in parallel - triage or take an explicit issue list, group by file-conflict, create a worktree per lane, delegate to sub-agents, open a PR per lane; never merges. Use when asked to implement issues, work the backlog, or pick up multiple issues in parallel. (skills/implement-issues/SKILL.md)
• issue: Create GitHub issues with proper formatting, labels, and AI implementation prompts. Use when asked to create an issue, report a bug, or request a feature. (skills/issue/SKILL.md)
• jira: Generate Jira-style ticket descriptions. Use when the user says "/jira", "jira ticket", or "give me a Jira ticket". (skills/jira/SKILL.md)
• lint: Run linting and formatting. Prefer uv run lintro chk for checks and uv run lintro fmt for formatting when lintro is available; fall back to native tools only when lintro is unavailable or another skill documents a follow-up pass (e.g. raycast). (skills/lint/SKILL.md)
• lintro-add: Guide for adding new linting/formatting tools to lintro. Use when implementing shellcheck, shfmt, sqlfluff, taplo, semgrep, gitleaks, or any new tool plugin. (skills/lintro-add/SKILL.md)
• lintro-review: AI branch review using lintro's own lintro review — a standard local pre-push pass alongside the coderabbit and greptile CLIs, and the review of record when those are rate-limited or unavailable. Use when asked for a lintro review, an interim/fallback AI review, or as part of the default pre-push review set. (skills/lintro-review/SKILL.md)
• lintro-verify: Verify that a lintro tool implementation is complete and follows all project standards. Use after adding a new tool to lintro. (skills/lintro-verify/SKILL.md)
• pr: Create pull requests with proper templates and metadata. Use when asked to create a PR, open a pull request, or submit changes for review. Auto-assign and auto-labeling handled by CI. (skills/pr/SKILL.md)
• pr-raycast: Prepare and open a pull request to raycast/extensions. Use when the user asks to open, submit, or get ready for a Raycast Store extension PR — not for general extension coding. (skills/pr-raycast/SKILL.md)
• properize: Promote a quick-and-dirty prototype to an lgtm-hq-standard repo - commit WIP lint-clean, grill the design, spec the backlog as milestone/epic/issue tree, then implement issue by issue. Use when asked to properize, productionize, or turn a prototype into a proper project. (skills/properize/SKILL.md)
• raycast: Raycast extension development standards. Use when writing or modifying Raycast extensions. Run lintro first, then Raycast's toolchain (npm run lint) which takes precedence for extension-specific rules. (skills/raycast/SKILL.md)
• rebase: Rebase the current branch onto the latest main. Use when asked to rebase, sync with main, update branch, or pull latest changes. Fetches and rebases onto latest main; does not push automatically. (skills/rebase/SKILL.md)
• reconcile: Consolidate worktrees and clean up stale branches. Use when asked to reconcile, clean up worktrees, consolidate branches, or tidy up a project's git state. (skills/reconcile/SKILL.md)
• scorecard: Audit the OpenSSF Scorecard rating for py-lintro. Use when asked to check the scorecard, understand the rating, or find what's missing. Specific to github.com/lgtm-hq/py-lintro. (skills/scorecard/SKILL.md)
• stand-ci: CI/CD and GitHub Actions guidelines. Use when writing workflows or Actions. Shell script code must be in dedicated .sh or .py files. Actions must be pinned to SHAs, not versions. (skills/stand-ci/SKILL.md)
• stand-general: Global coding standards for all projects and languages. Use when writing any code. Covers linting with lintro, testing with coverage, semantic commits, PR creation, and pre-push AI review with coderabbit, greptile, and lintro review CLIs. (skills/stand-general/SKILL.md)
• stand-odin: Odin coding standards. Use when writing Odin code. Covers idiomatic error handling with or_else and or_return, memory/allocator patterns, attributes, naming and API conventions, and testing with core:testing. (skills/stand-odin/SKILL.md)
• stand-py: Python >= 3.11 coding standards. Use when writing Python code. Requires type hints, return types, Google-style docstrings, trailing commas, explicit kwargs, StrEnum with auto(), dataclasses, pytest-style tests. (skills/stand-py/SKILL.md)
• stand-rust: Rust coding standards. Use when writing Rust code. Covers edition, error handling with thiserror/anyhow, unsafe policy, type patterns, testing, documentation, and dependency management. (skills/stand-rust/SKILL.md)
• stand-ts: TypeScript and JavaScript standards. Use when writing TS/JS code. Covers strict mode, type patterns, error handling, imports, naming, testing, React conventions, and package management with bun. (skills/stand-ts/SKILL.md)
• sweep-prs: Retrospective audit of recently merged/closed PRs for anomalies that slipped through - merges without green builds, merges onto red main, unresolved or post-merge review threads, hollow resolutions. Read-only sweep, then confirmed remediation (issues, disposition replies). Use when asked to sweep PRs, audit merged PRs, or check what slipped through. (skills/sweep-prs/SKILL.md)
• test: Run tests with coverage reporting. Auto-detects test frameworks (Vitest, Playwright, RSpec, pytest, Jest, BATS, etc.) and runs appropriate test commands. Use when asked to run tests, check coverage, or validate code. (skills/test/SKILL.md)
• test-api: Playwright API testing best practices. Use when writing REST API tests with Playwright. Enforces Zod schema validation, client/fixture separation, and contract testing. (skills/test-api/SKILL.md)
• test-shell: BATS shell script testing. Use when writing or running shell script tests. Covers setup/teardown, assertions, mocking, helper patterns, and coverage with kcov. (skills/test-shell/SKILL.md)
• test-ui: Playwright E2E testing best practices. Use when writing browser tests, visual regression, or accessibility tests in any project. Enforces user-facing locators, auto-waiting, web-first assertions, and Page Object Model. (skills/test-ui/SKILL.md)
• test-ui-qsf: QSF project conventions for the Playwright E2E suite. Use when working in the QSF playwright-tests directory (authFixtures, ___ping-* handles, data-test-label test IDs, .auth storage states) — not for generic Playwright advice. (skills/test-ui-qsf/SKILL.md)
• turbo-add: Guide for adding a new theme family to turbo-themes. Use when implementing Nord, Solarized, Gruvbox, Tokyo Night, One Dark, Ayu, Kanagawa, Everforest, Radix, or any new theme. (skills/turbo-add/SKILL.md)
• turbo-test: Run the full turbo-themes build and test pipeline. Use when asked to build, test, lint, or validate the turbo-themes project. Includes example projects by default. (skills/turbo-test/SKILL.md)
• turbo-verify: Verify that a theme implementation is complete and follows all project standards. Use after adding a new theme to turbo-themes. (skills/turbo-verify/SKILL.md)
• which-pr: Report which PR is being worked on in the current conversation. Use when asked about the current PR context. (skills/which-pr/SKILL.md)

Total skills: 40

Cursor Cloud specific instructions

There are no long-running services. This repo ships two deliverables: the Python catalog/validation/generator scripts (managed with uv) and the TypeScript gateway CLI (@lgtm-hq/ai-skills, managed with bun) under npm/ai-skills/.

• Toolchains: Python via uv (requires-python >= 3.13; uv provisions the interpreter, so the base image's Python does not matter) and JS via bun for the gateway package. The startup update script installs both and runs uv sync --frozen plus bun install in npm/ai-skills/.
• PATH: uv lands in ~/.local/bin and bun in ~/.bun/bin (the bun installer appends this to ~/.bashrc). If uv/bun are not found in a fresh non-login shell, prefix commands with those paths or export PATH="$HOME/.local/bin:$HOME/.bun/bin:$PATH".
• Standard commands live in CONTRIBUTING.md (root: uv run lintro chk, uv run lintro fmt, uv run lintro tst, bash scripts/validate.sh) and in npm/ai-skills/package.json (bun test). Run Python tests directly with uv run pytest tests/. scripts/validate.sh needs both uv and rg on PATH.
• Running the gateway CLI without installing: bun npm/ai-skills/bin/skill.ts <list|vendors|install|update|remove|adopt|doctor>.
• Gateway install/update shell out to the upstream Vercel skills CLI via bunx skills@^…, which needs network + GitHub access. To smoke-test a real install, prefer a vendor source (for example --vendor mattpocock --skill code-review -a cursor --project -y inside a scratch directory). First-party installs resolve lgtm-hq/ai-skills@v<package-version>, and on main the package version is 0.0.0-dev (not a real tag), so a first-party install will not resolve there.