Refine a draft task specification into a fully planned, implementation-ready task with acceptance criteria, architecture, per-step sub-task files and verifiable phases

Implement a task step by step with automated LLM-as-Judge verification at the end of each phase

Use when creating or developing, before writing code or implementation plans - refines rough ideas into fully-formed designs through collaborative questioning, alternative exploration, and incremental validation. Don't use during clear 'mechanical' processes

creates draft task file in .specs/tasks/draft/ with original user intent

Execute one complex task as ordered, dependent steps run sequentially, passing context from each step to the next, with per-step LLM-as-a-judge verification. Use when later steps depend on the results of earlier ones.

Run independent tasks concurrently across multiple files or targets using parallel sub-agents, with per-task model selection and LLM-as-a-judge verification. Use when tasks do not depend on each other and can run side by side.

Execute tasks through competitive multi-agent generation, meta-judge evaluation specification, multi-judge evaluation, and evidence-based synthesis

Execute a task with sub-agent implementation and LLM-as-a-judge verification with automatic retry loop

Review an existing GitHub pull request and post inline review comments on its diff. Use when the changes are on an opened PR rather than your local working tree.

Review your local uncommitted working-tree changes (git diff plus untracked files) and return actionable improvement suggestions. Use before committing, when nothing has been pushed yet.

Use when creating or editing any prompt (commands, hooks, skills, subagent instructions) to verify it produces desired behavior - applies RED-GREEN-REFACTOR cycle to prompt engineering using subagents for isolated testing

Workflow required before any Mule flow and integration work. Call use_skill as your FIRST action — before reading project files — whenever the user asks to create, generate, update, fix, modify, change, edit, tweak, adjust, or rework any Mule flow, sub-flow, or component. Do not read project files and attempt the change yourself — even targeted single-component changes like 'modify the choice router', 'fix the until-successful', or 'update the catch block' require this workflow. Covers all change types, new integrations and targeted changes to error handlers, catch blocks, choice routers, DataWeave transforms, HTTP listeners, foreach loops, retry policies, scatter-gathers, connectors, and variable assignments. Prompts beginning with 'This code defines...' or 'This flow...' are generation requests, not analysis. When you call this skill, it must be the only tool call in that response.

Use when bumping Meteor package versions for beta, RC, or official releases. Covers the two version schemes (meteor-tool vs all other packages), the update-semver automation tool, manual files the tool does not handle, modern-tool npm package semver analysis, the official installer gate, and the full lifecycle from beta through official release. Applies to packages/*/package.js, scripts/admin/, npm-packages/meteor-installer/, modern-tool npm packages, and .meteor/versions in test apps.

Use when writing tests, debugging test failures, running the test suite, or setting up test infrastructure. Covers self-test, package tests, and modern E2E tests.

Generate Django database migrations for Sentry. Use when creating migrations, adding/removing columns or tables, adding indexes, or resolving migration conflicts.

Guide for using Sentry's layout and text primitives. Use when implementing UI components, layouts, or typography. Enforces use of core components over styled components.

Reference and active migration guide for Sentry's cell architecture. Explains what cells and localities are and why they're different, how requests reach cells via Synapse API routing, ingestion routing, and the control silo gateway, and how to safely query cross-cell data without silently missing results. The migration section covers how to do migration work: draining the URL_NAME_TO_ACTION registry in test_urls.py to zero (with a recipe for each action type), rolling deploy safety and the two-phase pattern required by independent sentry/getsentry deploys, and the region -> cell rename including what not to rename (DB columns, AWS refs, uptime regions, billing address). Also documents known issues with proposed fixes: integration TeamLinkageView routing, Jira cross-cell fan-out, and relocation endpoint routing.

Instrument and discover analytics events in Sentry's frontend UI. Use when adding tracking to buttons, pages, modals, or custom interactions, when defining new analytics events, when searching for existing events, when auditing analytics coverage for a feature, or when answering questions about how users interact with a feature. Trigger on "add analytics", "track event", "instrument analytics", "analytics event", "track click", "track page view", "add tracking", "what events exist for", "audit analytics", "how many people", "how many users", "are people using", "is anyone clicking", "usage of", "who is using".

FastAPI best practices and conventions. Use when working with FastAPI APIs, Pydantic models, dependencies, streaming responses including Server-Sent Events (SSE), and serving frontend apps. Keeps FastAPI code clean and up to date with the latest features and patterns.

Create a git worktree with a new branch for implementing a fix or feature for a GitHub issue