ai-kit

Create conventional git commits from all uncommitted changes. Extracts ticket numbers from branch names and uses conventional commit format.

Create conventional git commits from staged changes only. Extracts ticket numbers from branch names and uses conventional commit format.

Push current branch and create a GitHub pull request with proper formatting. Use when you have committed changes ready to be pushed and want to open a PR with a well-structured description.

Analyze and fix GitHub issues. Use when given a GitHub issue URL or number to investigate, understand the root cause, implement a fix, and verify the solution works.

Interview users in-depth to create detailed feature specifications. Use when gathering requirements for a new feature, understanding user needs, or creating PRDs. Asks probing questions about technical implementation, UI/UX, and edge cases.

Generate and maintain project documentation. Use when initializing docs for a new project, updating architecture docs, or creating feature specifications. Creates ai_docs/ and ai_specs/ folders with structured documentation.

Update implementation plan files by marking completed tasks and committing changes. Use after completing tasks in a phased implementation to track progress and keep the plan file current.

Review code against plan and standards using code-reviewer agent

Full commit and push workflow with validation — groups files into logical commits

Create or update HANDOFF.md for context continuity between sessions

Interview to flesh out a plan/spec

Postgres performance optimization and best practices from Supabase. Use this skill when writing, reviewing, or optimizing Postgres queries, schema designs, or database configurations.