laurigates/claude-plugins

zsh-gotchas

Four zsh expansions that silently rewrite a command: $VAR:word modifiers, a leading =, no word splitting, `path` tied to PATH. Use when a zsh one-liner gives a wrong result or a tool seems missing.

evaluate-plugin-batch

Batch evaluate every skill in a plugin and produce a plugin-level report. Use when auditing an entire plugin's quality or validating before a release.

evaluate-skill

Evaluate a skill by running test cases and grading results. Use when testing whether a skill produces correct guidance, validating improvements, or benchmarking before release.

git-issue

GitHub issue to PR end-to-end — branch, TDD implementation, PR — one issue or several in parallel. Use when asked to work on an issue, fix issue #N, or batch-process several.

git-pr-feedback

Address PR review comments and resolve threads. Use when CHANGES_REQUESTED is set, working through unresolved review threads, or replying to reviewer feedback.

test-analyze

Analyze test results and create a fix plan with subagents. Use when triaging failing tests, analyzing JUnit XML, planning fixes for accessibility/security, or categorizing flaky/E2E failures.

parallel-agent-dispatch

Dispatch contract for spawning parallel agents covering worktree collisions, scope overflow, and silent exits. Use when fanning out concurrent agents or authoring a lead prompt.

health-check

Claude Code health check — scans plugins, settings, hooks, MCP, runtime state, usage telemetry, permissions, marketplace with optional fixes. Use when checking project health or troubleshooting setup.

session-distill

Distill session insights into rules, skill improvements, recipes, and cross-repo promotions to marketplace plugins. Use when capturing learnings, codifying workflow into .claude/rules, or promoting a session-invented pattern into a specific plugin/skill as a…

workflow-interrupted-run-recovery

Recover a Workflow run killed mid-flight: salvage killed agents' worktrees, resume without duplicating PRs already opened. Use when a run reports agents_error after a session limit, kill, or crash.

plugin-registry

Claude Code plugin registry structure, install scopes, and version lag. Use when troubleshooting plugin problems, a stale plugin version in one project, or fixing registry entries.

mcp-management

Install, configure and troubleshoot MCP servers. Use when adding/enabling servers, editing .mcp.json, fixing OAuth, or when a server runs stale code after an upstream fix.

session-end

End-of-session orchestrator. Previews which of wrap/distill/feedback/taskwarrior-sync qualify, single confirm, then sequence. Use when winding down a session.

session-spinup

Read-only session-start briefing of open tasks, git state, journal todos — reports, never edits code. Use when user says spin up, what was I doing, or pick up where I left off.

session-wrap

End-of-session capture to taskwarrior, optional journal, GitHub issues. Use when user says wrap up, session wrap, or done for now.

upstream-pr-contribution

Opening a pull request against a repository you do not own. Use when about to open one, to read CONTRIBUTING.md and the PR template in full and meet the contribution gates bots enforce.

worktree-stale-base-merge

Merging a PR from a parallel worktree-agent branch, or any branch cut from an older main. Use when deciding whether a green CI check still means anything after sibling branches have landed.

ai-review-max-turns

A Claude-powered CI review check reports red. Use when triaging it, to separate a genuine finding the check could not publish from turn-budget exhaustion — read is_error first, never the red X.

multirepo-ci-cd

Diagnose a CI failure or roll a workflow out across many repos. Use when debugging a red check in a multi-repo org, promoting a shared reusable-*.yml, or re-triggering a PR after a @main fix.

tool-migration-cutover

Migrating between tools (Dependabot to Renovate, PAT to GitHub App, one linter to another). Use when removing the incumbent, to verify the replacement actually runs rather than merely configured.

git-security-checks

Pre-commit security validation and secret detection via gitleaks. Use when scanning for secrets, setting up gitleaks, or configuring .gitleaks.toml pre-commit security.

multi-repo-discipline

Multi-repo workspace rules: read-only fixtures, upstream/downstream pairs. Use when dispatching agents across sibling repos or editing another repo's .claude/.

configure-justfile

Check and configure Justfiles with standard recipes. Use when setting up a Justfile, auditing for missing recipes, or migrating from Makefile to Justfile.

justfile-expert

Just command runner expertise — Justfile syntax, recipes, parameters, modules, shebang recipes. Use when authoring justfiles, project commands, or task automation.

git-issue-scoping

Read an issue's full comment thread and re-verify its cited evidence at HEAD. Use when scoping a PR or plan from a GitHub issue, or before filing or reversing one in your own tracker.

project-test-loop

Test-fix-refactor loop that edits code between runs until the suite is green, bounded by --max-cycles. Use when asked to make failing tests pass or drive a TDD RED/GREEN/REFACTOR cycle.

test-focus

One test file, fail-fast — Playwright, Vitest, Jest, pytest, cargo, go test. Use when running a single spec file, headed/debug mode, or serial single-worker WebGL tests.

test-full

Every tier in pyramid order — unit, then integration, then E2E. Use when running all tests before a PR or commit, or generating a coverage report.

test-quick

Unit tier only — skips slow, integration, and E2E tests for sub-30s feedback. Use when checking units after a change, running tests affected since the last commit, or unit watch mode.

test-report

Cached results from the last test run, no re-run — pass/fail counts, existing coverage report, history, flaky tests. Use when asked for current test status or standup health.

test-run

Run the project's tests — auto-detects pytest, vitest, jest, cargo, go test; whole suite or a name pattern. Use when asked to run the tests and no tier, single file, or fix cycle is named.

test-tier-selection

Decide which test tier a change needs — unit, integration, or E2E — from its scope; runs nothing itself. Use when deciding how much of the suite to run after a change.

multi-model-delegation

Multi-model design consults via PAL (kimi, glm, gemini, gpt). Use when asking other models to brainstorm a design or reconciling their split answers.

configure-security

Security scanning: dependency automation, SAST, secrets detection. Use when setting up Renovate/Dependabot, CodeQL, or TruffleHog in CI, or creating a SECURITY.md policy.

git-merge-hazards

Traps in GitHub's merge machinery. Use when merging a PR, merging a stacked PR chain, auditing whether a branch really landed, or merging over red CI.

git-pr

Create pull requests with descriptions, labels, and issue references. Use when user says "create PR", "open pull request", or "submit for review". From pushed branches.

git-rebase-patterns

Advanced git rebase — linear history, stacked PRs, --update-refs, --onto. Use when rebasing branches, cleaning history, managing PR stacks, or fixing merge-heavy branches.

git-stacked-prs

GitHub native stacked PRs via gh stack. Use when creating, syncing, rebasing or merging a PR stack, or choosing stacks vs manual PR chains.

mutation-testing

Mutation testing with Stryker (TS/JS) and mutmut (Python). Use when finding weak tests that pass on mutated code, or improving test quality through mutation analysis.

evaluate-context-engineering

Context-engineering audit (C1-C6) of skills and always-loaded rules. Use when trimming context bloat, auditing a plugin, or after a new Claude model ships.