Analyze Repo
Full repository analysis. Invoke at the start of a new session or when Claude needs to understand the project from scratch. Writes a comprehensive summary to SESSION_STATE.md.| Use when: starting fresh, after a long break, when context was lost, or when explicitly asked to "analyze the project".
Git Workflow
Git workflow assistant. Invoke when the user wants to commit, create a branch, write a PR description, or follow the project's git conventions. Auto-activates on: "commit this", "create a PR", "write a commit message", "what should I name this branch".
Code Review
Structured code review. Invoke before committing, before opening a PR, or when you want feedback on changes. Auto-activates on "review this code", "check my changes", "look for issues", "review before I push". Routes to code-reviewer sub-agent to keep main context clean.
Doctor
Self-heals this project's Claude Code setup. Diagnoses broken or missing pieces — skills with no SKILL.md, hooks referenced but absent or non-executable, invalid settings.json, bloated CLAUDE.md — scores health, and applies safe fixes. Use when something in the setup seems off, after edits to .claude/, when a hook or skill isn't firing, or on a `[heal]` nudge. Triggers: 'is my setup ok', 'fix my…
Learn
Makes the setup remember. Captures durable, project-specific learnings — user preferences, stack facts, conventions, fixes, recurring gaps — so every future session starts smarter (they're injected at SessionStart). Use when you discover something about how THIS project or user works that should persist. Triggers: 'remember that…', 'from now on…', 'we always/never…', or right after you learn a no…
Augment
Gives this project a NEW capability on demand. Use the moment the user needs something the current setup can't do well — a new framework, a service integration (Stripe, Postgres, Twilio…), a deploy target, a niche workflow. Searches the open ecosystem for a fitting skill/MCP across platforms, installs the best vetted one, and if nothing fits, forges a bespoke skill. Triggers: 'can you also…', 'I…
Onboarding
First-run user onboarding. Asks 6 quick questions, detects the project, writes USER_PROFILE.json, and recommends the right setup tier. Auto-activates on first use or when the user says 'set me up', 'I'm new', or runs /onboard.
Tips
Surfaces a few high-leverage, context-aware tips for getting more out of Claude Code. Self-contained (no external data file). Invoked by /tips, by analyze-repo at the end of a scan, or when the user asks how to use Claude better.
Test Runner
Run the project's test suite and get a structured report. Invoke when you want to run tests after making changes. Auto-activates when user says "run tests", "do tests pass", or "check if this broke anything". Routes execution to the test-runner sub-agent to keep main context clean.
Self Update
Checks whether the project's claude-bootstrap setup is current and applies updates safely. Compares the installed version against the source repo, shows what changed, and updates only what the user approves. Triggered by /update.
Security Scan
Security audit skill. Invoke before deploying, after adding auth code, or when reviewing for vulnerabilities. Auto-activates on "security check", "check for vulnerabilities", "pre-deploy audit", "scan for secrets", "is this auth secure". Routes to security-scanner sub-agent.
Forge
Authors a NEW, project-specific Claude Code skill from scratch when the ecosystem has nothing that fits. Use after `augment` finds no good match, or when the user wants a custom workflow captured as a reusable skill. Produces a detailed, validated skill — never a generic stub. Triggers: 'make a skill for…', 'turn this into a skill', 'I keep doing X, automate it'.
Dep Check
Dependency audit skill. Check for outdated packages and security vulnerabilities. Auto-activates on "check dependencies", "any vulnerable packages", "audit deps", "what's outdated", "npm audit". Routes to dep-checker sub-agent to keep main context clean.
Context Guard
Context window protection. Auto-activates when the conversation is getting long, when the user is about to do a large file read or exploration task, or when the context is approaching its limit. Helps route expensive operations to sub-agents to protect the main session.