Share your AI coding setup — CLAUDE.md, skills, cursorrules — with one command. install / preview / rollback.
Agent Claude Code
Run mutation testing to verify tests actually catch broken behavior, not just execute lines. Detects shallow suites where coverage looks healthy but assertions are missing. Use after major test changes, before declaring a suite production-ready, or when bugs slip through despite green CI. Installs the mutation…
Agent Claude Code
Flag code that is heavier than its problem — single-implementation abstractions, speculative generalization, unnecessary indirection, unused config, premature optimization, and type gymnastics for states that can't occur. Scope-first, read-only, proposes the simpler alternative with its cost. Use in PR review, before…
Agent Claude Code
Performance-marketing specialist for paid acquisition on Meta (Facebook/Instagram) and Google Ads. Plans, audits, and optimizes campaigns — structure, targeting, budget pacing, creative testing, bidding, and incrementality — using the meta-ads and google-ads MCP tools. Use for ad-account audits, campaign structure…
Agent Claude Code
Execute implementation plans by dispatching a fresh subagent per task with mandatory two-stage review (spec compliance then code quality) after each task. Use when you have a written plan with mostly independent tasks and want high-quality same-session execution without context pollution between tasks. Enforces…
Agent Claude Code
Execute phased plans by fanning out one agent per task per wave, reconciling per wave, gating between phases with verify commands, and emitting a phase × outcome report. Use for "execute this plan", "work through these phases", "swarm over this backlog" — any plan with ≥3 total tasks or ≥2 tasks in a single phase that…
Agent Claude Code
Strategic planning consultant with interview workflow (Opus).
Agent Claude Code
Two-axis review of a git diff — Standards (does the code follow this repo's documented conventions?) and Spec (does it match what the issue/PRD asked for?). Runs both axes as parallel sub-agents and reports them side by side without merging findings. Use when reviewing a branch, PR, or work-in-progress changes.
Agent Claude Code
Interactive CLI testing specialist using tmux for session management.
Agent Claude Code
Run retrieval regression gates (hitgate) against the current repo state. Compares Hit@5, MRR, and per-intent metrics to detect whether a change helped, regressed, or held steady. Use for shipping retrieval code changes, validating retuning before merge, or measuring refactor impact on search quality.
Agent Claude Code
Orchestrate end-to-end refactors across 6 phases: RAG pre-flight (prior context + protected scopes), plan with rollback, critic scope gate, parallel 3-agent execution (architect/builder/reviewer), two-stage review (spec+quality), test cleanup, ADR capture, sync. Use for scope >5 files, cross-module boundaries, or…
Agent Claude Code
Release engineering and CI/CD specialist for GitHub Actions pipelines, release-please automation, semantic versioning, changelogs, branch protection, flaky-test diagnosis, and runner cost optimization. Use for: CI failures and diagnostics, release-please gate issues, version bumps and changelogs…
Agent Claude Code
Executes the /repaint pipeline — register lock, reference anchor, token spec (or DESIGN.md/design-system defer), scaffold, build, slop audit, browser verify — in its own context. Dispatched by the /repaint skill, 1× or N-parallel (worktrees). Builds and verifies; never deploys. (Sonnet).
Agent Claude Code
Evaluate library, pattern, or architecture choices end-to-end: research candidates, challenge with decision-critic, plan adoption, write ADR with revisit-when condition, index for future recall. Use for any choice where the wrong decision creates technical debt or lock-in. Always produces a durable ADR.
Agent Claude Code
Data analysis and research execution specialist.
Agent Claude Code
Security vulnerability detection specialist (OWASP Top 10, secrets, unsafe patterns).
Agent Claude Code
Review skill-creator benchmark outputs for a completed iteration. Reads benchmark.json, grading.json, evalmetadata.json, and response.md files from an iteration directory, then surfaces discriminating assertions, anti-patterns in winning outputs, and improvement recommendations. Use after skill-creator eval runs…
Agent Claude Code
Apply 4-phase systematic debugging to any bug, test failure, or unexpected behavior. Enforces root-cause investigation before proposing any fix, tracks turn efficiency (file read budget, edit budget, subagent escalation), and blocks rationalization attempts. Use when encountering any technical failure — especially…
Agent Claude Code
Enforce test-driven development discipline for features, bug fixes, and refactors. Writes failing tests first, watches them fail, implements minimal code, refactors under green. Use whenever writing production code — blocks implementation until a failing test exists. Harder TDD discipline than test-engineer (which…
Agent Claude Code
Decompose a task into parallel workstreams, assign agent ownership, run integration at dependency boundaries, and synthesize results. Use when a task is large enough that parallel agents save time or add confidence — and when clear handoffs can be defined. Produces a team plan, bounded prompts per agent, and final…
Agent Claude Code
Test strategy, integration/e2e coverage, flaky test hardening, TDD workflows.
Agent Claude Code
Evidence-driven causal tracing with competing hypotheses, evidence for/against, uncertainty tracking, and next-probe recommendations.
Agent Claude Code
Video editing craft expert for story, pacing, hooks, and retention. Directs editing strategy for short-form (reels/shorts) and long-form (YouTube) with emphasis on hook engineering, cut timing, sound design, captions, and retention curves. Knows the shorts-edit CLI for batch programmatic editing. Use for edit…
Agent Claude Code
Full-stack web application developer for React/Next.js + Supabase. Use when building features that span frontend components, Supabase schema/auth/realtime, and API integration. Covers schema design, RLS-aware data access, auth flows, realtime subscriptions, and production deployment.
Agent Claude Code
Technical documentation writer for README, API docs, and comments (Haiku).
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: