matrummet

Kill all running local development processes (TanStack Start dev, email service, PostgREST), wait for termination, then restart them. Use when user says "boot local", "restart local", "reboot dev", or needs to refresh local services.

Run user story acceptance tests using Playwright browser automation. Executes test steps from user-stories/*.md files and reports pass/fail status with reasons. Use when user says "run user stories", "test user stories", "acceptance test", or "run e2e tests".

Scrape, load, and manage the base recipe pool for the meal planner. Handles scraping Swedish recipe sites, loading into database, and verifying integration. Use when user says "scrape recipes", "load base recipes", "base recipe pool", or "update base recipes".

Evaluate test coverage and create appropriate tests based on risk assessment. Use when user says "build tests", "update tests", "add test coverage", "write tests for this", or "improve test coverage".

Reviews code quality, pattern adherence, and architecture with red-team validation. Use when user says "check your code", "code quality review", "is this well-written", or "review code quality". Different from check-your-work which finds bugs.

Orchestrates 6 quality agents in parallel (duplicate-detector, user-intent, deep-bug-hunter, security, performance, correctness) with severity validation. Use when user says "check your work", "review what we wrote", "quality check", or after writing new features.

Create distinctive, production-grade frontend interfaces with high design quality. Use when user asks to build web components, pages, or applications. Generates creative, polished code that avoids generic AI aesthetics.

Update dependencies in a workspace (default: frontend). Shows outdated packages, updates them, runs tests to verify. Use when user says "update deps", "update dependencies", "upgrade packages", or "update frontend deps".

Validates AI implementation plans before execution. Use when user says "check your plan", "validate this plan", "review the plan", or "is this plan good". Launches 5 parallel validators + devil's advocate.

Remove signs of AI-generated writing from text. Use when editing or reviewing text to make it sound more natural and human-written. Based on Wikipedia's comprehensive "Signs of AI writing" guide. Detects and fixes patterns including: inflated symbolism, promotional language, superficial -ing analyses, vague attributions, em dash overuse, rule of three, AI vocabulary words, negative parallelisms, and excessive conjunctive phrases.

Full implementation pipeline from idea to commit. Explores codebase, plans, validates plan, implements, reviews work, reviews code, tests, commits, and validates with user story acceptance tests. Use when user says "ship it", "full pipeline", "end to end", or wants complete implementation workflow.

Convert a feature request into a user story, save it, implement it via ship-it, and validate with acceptance tests. Use when user says "new user story", "implement this story", "story to ship", or describes a feature they want built as a user story.

Run all tests (vitest), lint (eslint), and commit changes. Use when user says "test and commit", "run tests and commit", "verify and commit", or "CI and commit".

Diagnose failing tests to determine if failure is code bug or test issue. Use when user says "test or bug", "is this a test bug", "why is this test failing", or "diagnose test failure".

Transform documentation into AI-optimized format for Claude Code. Use when user says "update doc", "optimize doc", "rewrite for claude", or "AI optimize". Shows recently modified .md files for selection, verifies accuracy against codebase, generates optimized version, compares for completeness, then replaces with approval.