Workflows

Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skill should be used when the user is looking for functionality that might exist as an installable skill.

Refines a v1 project plan into agent-ready tasks with clear context, implementation steps, and validation criteria. Use after /plan-project has produced a v1 plan.

ESLint, Prettier, Ruff, mypy, and TypeScript strict mode configuration patterns. Use when setting up linting, formatting, or code quality tooling for a new project, reviewing existing configs, generating config files, or adding lint steps to CI. Triggers on tasks involving ESLint, Prettier, Ruff, mypy, Black, code style, linting, formatting, or code quality enforcement.

Automates browser interactions for web testing, form filling, screenshots, and data extraction. Use when the user needs to navigate websites, interact with web pages, fill forms, take screenshots, test web applications, or extract information from web pages.

Socratic discovery and design exploration before planning. Activates when objective complexity criteria are met (2+ modules, 4+ tasks, 2+ approaches, or new patterns/integrations) — asks clarifying questions, explores alternatives and tradeoffs, produces a design document for approval. Pulls context from Linear issue description, linked docs, and existing CLAUDE.md learnings.

Creates Linear issues from a refined project plan. Validates issues were created correctly with sufficient context for agent execution. Use after /refine-plan.

Creates an isolated git worktree for task execution. Activates when starting work on a planned issue — sets up a new branch with Linear issue ID, runs project setup, and verifies a clean test baseline before coding begins.

After a PR is opened (or on demand), read Greptile's 0–5 confidence score and converge the PR toward 5/5 — grill on intent, fix review findings, re-trigger Greptile, loop up to 3 rounds, then run a final independent review. Use during /workflows:ship right after the PR is created, or when the user asks to check the Greptile score, run the Greptile gate, or read/address the Greptile review on a PR. Skips gracefully when Greptile isn't on the repo.

Breaks work into bite-sized tasks before coding. Activates when a multi-step task needs planning — creates tasks small enough for a junior developer to follow (2-5 minutes each), with exact file paths, complete implementation details, and verification steps. References Linear issue context and project-specific test commands from CLAUDE.md.

Executes a structured plan using subagent-per-task with TDD enforcement. Activates when given an approved plan to implement — launches fresh subagents for each task, enforces red-green-refactor, runs two-stage review per task, and checkpoints between tasks. Parallelizes independent tasks.

Runs the full post-plan workflow. Refine plan, create Linear issues, setup CLAUDE.md. Use after /plan-project produces a v1 plan. Pauses between phases for optional review.

Generates a best-practices CLAUDE.md file for the project. Analyzes the codebase and applies Claude Code best practices for optimal agent performance. Use at project setup or after /create-issues.

Searches project-level precedent INDEX files for relevant past decisions. Activates during brainstorming or planning when historical decisions may inform the current approach — searches by keyword against Decision and Tags columns, filters by category, lazy-loads full trace files on match. (Org-level handbook search is currently unavailable.)

Use when writing, reviewing, or refactoring test code. Triggers on Vitest tests, React Testing Library usage, MSW handlers, Playwright E2E tests, pytest fixtures, test structure decisions, or coverage configuration. Contains 46 testing rules across 12 categories.

Captures durable knowledge after completing work so future sessions benefit. Activates during the ship phase — updates CLAUDE.md with architectural decisions and gotchas, writes session summary to auto-memory, updates docs if architecture or API changed. Only records genuinely durable facts, not session-specific noise.

Audits and auto-fixes a project's CLAUDE.md against Anthropic best practices. Activates during ship phase — checks conciseness, enforces @import structure for detailed docs, auto-excludes bloat, identifies hook candidates, and auto-fixes structural issues. Flags content questions for developer review.

Detects handbook content that has drifted from project reality after shipping. Activates during the ship phase — compares the shipped diff against handbook files via GitHub API, identifies stale or missing handbook content, and optionally opens a PR to the handbook repo with proposed updates.

React and Next.js performance optimization guidelines from Vercel Engineering. This skill should be used when writing, reviewing, or refactoring React/Next.js code to ensure optimal performance patterns. Triggers on tasks involving React components, Next.js pages, data fetching, bundle optimization, or performance improvements.

Four-phase root cause analysis for bug investigation. Activates when debugging unexpected behavior, failing tests, or production issues — follows reproduce, isolate, analyze, fix with defense-in-depth. Uses condition-based waiting instead of arbitrary delays. Available anytime, not tied to the inner loop sequence.

Use when writing, reviewing, or refactoring FastAPI/Python backend code. Triggers on FastAPI endpoints, Pydantic models, SQLAlchemy queries, async Python code, or Python API architecture. Contains 38 architectural rules across 8 categories.

Ensures tasks are genuinely resolved before marking them done. Activates at task checkpoints during plan execution — validates that fixes actually work, tests genuinely pass, and acceptance criteria are met. Prevents premature completion declarations.