19 AI agents in your git workflow. Catches secrets, injection flaws, N+1 queries, and bad migrations on every commit. Structured Research → Plan → Implement workflow with GO/NO-GO gates before you write a line. Generates production-ready code, product wikis, and codebase maps on demand. Any language. Free.
Agent Claude Code
Generates docs/BLUEPRINT.md — a living map of the project covering stack, architecture diagram, API inventory, DB schema ER diagram, backend module map, and frontend component tree. Works on existing codebases, spec-only projects, and both together (with drift detection).
Agent Claude Code
Reviews code for quality issues including DRY violations, high cyclomatic complexity, poor naming, missing edge case handling, error handling gaps, dead code, and language-specific anti-patterns. Language-agnostic. Use on any code change.
Agent Claude Code
Writes production-complete feature implementations with enterprise best practices baked in — clean architecture layers, rate limiting, auth middleware, input validation, structured error handling, request logging, pagination, caching, and transactions. Goes beyond scaffolding — no TODOs, no placeholder logic.
Agent Claude Code
Validates features, technical decisions, and code changes against the project's CONSTITUTION.md principles — ethics, AI safety, human oversight, scope alignment, and architectural commitments. Runs in the RPI implement phase and standalone via /rpi-implement.
Agent Claude Code
Validates database migration safety before commit. Detects blocking operations on large tables, missing rollbacks, irreversible changes, unsafe NULL constraints, index creation without CONCURRENTLY, and breaking schema changes. Supports Prisma, Alembic, Django, Flyway, TypeORM, Sequelize, raw SQL, and Go Migrate. Only…
Agent Claude Code
Maintains project documentation. Creates README.md if missing, updates it when the project changes, appends to CHANGELOG.md with structured entries, and identifies code that needs inline documentation. Run after successful commits and on demand.
Agent Claude Code
Synthesizes outputs from multiple RPI research agents into structured artifact files — RESEARCH.md, pm.md, ux.md, eng.md, and PLAN.md. Also writes implementation records (IMPLEMENT.md). Called by RPI orchestrating commands.
Agent Claude Code
Detects performance issues in code changes: N+1 query patterns, missing database indexes, unnecessary re-renders, memory leaks, blocking operations in async contexts, inefficient algorithms, missing caching, and bundle size regressions. Use on backend, frontend, and data processing code changes.
Agent Claude Code
Generates comprehensive PR summaries after all automated tests pass. Analyzes the full diff against the base branch, describes what changed and why, identifies breaking changes, lists test coverage, and flags anything a reviewer should pay special attention to.
Agent Claude Code
Evaluates feature viability from a product perspective — user value, strategic alignment, market fit, and constitution compliance. Produces a viability score and recommendation for the RPI research phase.
Agent Claude Code
Generates concrete, ready-to-apply fix suggestions for issues found by review agents. Reads only the flagged files and lines — never re-scans the full codebase. Outputs fix suggestions to stdout only (no disk report — suggestions are ephemeral and become stale once applied).
Agent Claude Code
Analyzes a feature REQUEST.md and extracts structured requirements, goals, constraints, complexity estimate, and clarifying questions for downstream RPI planning agents. Invoked by /rpi-research.
Agent Claude Code
Synthesizes raw findings from review agents into the final review verdict. Owns the canonical implementation of .mantaignore suppression, inline manta-ignore checks, and cross-agent deduplication. Produces the machine-readable COMMITVERDICT/PUSHVERDICT blocks parsed by the git hooks (commit/push modes) or the full…
Agent Claude Code
Generates consistent boilerplate for new features by reading the minimal codebase context needed — one similar existing feature, its test, and the project config. Writes code files directly. No report generated — the output is the code itself. Optionally validates generated code against SPEC.md before writing.
Agent Claude Code
Full security audit agent. Checks for OWASP Top 10 vulnerabilities, hardcoded secrets, vulnerable dependencies, injection flaws, broken authentication, insecure data exposure, SSRF, path traversal, and more. Use on any code change — always runs in pre-commit pipeline.
Agent Claude Code
Assesses technical feasibility, designs implementation architecture, estimates effort, and writes the engineering plan (eng.md) and phased PLAN.md for RPI features. Also executes implementation tasks during /rpi-implement.
Agent Claude Code
Synthesizes product and technical research into a final GO/NO-GO recommendation. Weighs strategic value, technical risk, effort, and constitutional alignment to give a clear decision with rationale. Used in /rpi-research Phase 4.
Agent Claude Code
Converts design inputs (screenshots, Figma exports, wireframe images, or text descriptions) into production-ready UI components that match the project's existing design system, naming conventions, and framework. Detects component libraries (shadcn/ui, MUI, Chakra, Ant Design), icon libraries, dark mode strategy…
Agent Claude Code
Designs user experience flows, interaction patterns, and UI specifications for features in the RPI planning phase. Produces ux.md with wireframe descriptions, user journeys, and component specs.
Agent Claude Code
Generates a product wiki in docs/wiki/ — discovers all routes/screens by framework, captures screenshots when browser tools are available, analyzes feature code, compares to spec/SPEC.md when present, asks the user clarifying questions about ambiguous features, then writes structured markdown documentation.