AI Agent Workflow Collection

Add a new tool to the dotfiles repository. Use when asked to add a new CLI tool, application, or package as a managed dotfiles tool.

Draft a TIL-style blog post in the user's voice and write it to Notion with Status="Claude Draft". Contains voice guide for matching the user's writing style. Use when user approves a TIL topic and wants a draft created.

Fetches and processes GitHub pull requests waiting for review.

Provide convincing evidence that the code actually achieves its intended outcomes and does what it is expected to do runtime across all logical branches and edge cases. TRIGGER whenever the user asks you to prove, demonstrate, or provide hard evidence that code works as expected at runtime.

Look for opportunities to optimize the approach and/or implementation of the solution to a problem you see on the current branch (relative to its base branch). Use when the user asks you to assess, analyze, improve, optimize, review or validate the current branch or its changes.

Confirm a task was fully executed and the change works — static completeness audit followed by empirical runtime evidence. TRIGGER when asked to confirm a task is complete, prove it works, or verify a change is ready to ship.

Autonomously review and fix the current branch or specified scope in iterative rounds until clean, leaving all changes in the working tree. Escalates only genuine design decisions. Ends with a transparent report of every round, every fix, and anything it couldn't resolve — persisted to .agents/.

Fetch PR review comments, validate each against the actual code, auto-fix obvious issues, and escalate genuine design decisions. Leaves all changes in the working tree. Use when you want to action reviewer feedback autonomously.

Fetch and present a prioritized summary of all PRs awaiting the current user's review across the relevant projects and provide an interactive workflow helping the user phrase and post inline review comments and overall review decisions.

Review a PR, branch or file path and recommend the highest ROI improvement opportunities. TRIGGER when the user asks if the code is ready, looks good, or is otherwise complete.

Analyze gaps in user-facing features and desirable characteristics, developer-facing tooling and desirable characteristics, and other opportunities not currently…

Audit the user's Claude Code agent harness for improvements — global config, project config, recent session patterns, and opportunities for better guardrails and feedback loops. Use when the user asks to improve their Claude Code setup.

Find all tracked issues and work item ideas for the current project and rank them by priority to address. TRIGGER whenever the user asks for help deciding what to work on next and you are not already working through a known plan.

Identify the most mind-blowing, game-changing changes that could be made next assuming users will shortly abandon this project if it doesn't differentiate itself soon. TRIGGER whenever the user asks what work would be most impactful to do next.

Check whether a codebase upholds its invariants — a specific theme or all themes in parallel. Use when asked to review conventions, check invariants, or audit codebase health.

Scans GitHub commit history for commits that might make good TIL blog posts. Queries all your repos across all orgs via GitHub API. Tracks assessed commits in Notion to avoid duplicates across machines. Use when user asks for TIL ideas from their recent work.

[UNDER DEVELOPMENT - DO NOT USE] Searches the Notion Writing database for unpublished items that could become TIL posts. This skill is not yet integrated with the publishing workflow.

Load relevant code quality reference files (architecture, testing, security, type design, etc.) and apply their invariants to the current task. ALWAYS invoke before designing, writing, or editing ANY code.

Full reference for the beads task management workflow. Use when working with beads issues, planning multi-step work, or managing task lifecycle.

Reference for using Codefresh via the CodeFresh CLI. Use when a Codefresh CI check is failing in a recursionpharma PR, or when working with Codefresh builds.

Reference for using the NextJS Dev Tools via the nextjs-agent. Use to debug or observe Next.js framework internals.

Reference for using the Notion API via the notion-agent subagent. Use to expore and update Notion workspace content.

Full reference for the trekker task management workflow. Use when working with trekker issues, planning multi-step work, or managing task lifecycle.

Voice guide for writing PR descriptions. Use this skill for ALL PR creation and updates (via gh CLI or editing existing PRs). Contains your specific voice rules, anti-patterns, examples, and workflow. Never write PR descriptions without invoking this skill first.