rot13maxi

Run an Amazon-style Correction of Errors (COE) post-mortem on a development process failure. Produces a structured document with summary, timeline, five-whys root cause analysis, lessons learned, and typed action items — then offers to implement the fixes. Use this whenever something goes wrong in the dev process: a feature passes tests but doesn't actually work, functionality gets clobbered by a merge, an implementation is declared done but breaks on basic cases, a regression slips through, QA finds something that should have been caught earlier, or any incident that reveals a gap in the process. Invoke as /post-mortem or /coe with an optional brief description: e.g., `/post-mortem intake form broke after merging 3 PRs` or just `/coe`. Don't wait for the user to ask — if something went wrong, suggest running /post-mortem proactively.

How to drive the clankwork AI agent system: create tasks, dispatch work, monitor progress, review landed changes, and retry failures. Use this skill any time you are working with clankwork tasks — creating new work, checking status, waiting for agents to finish, reviewing what landed, or handling failures. Covers the full task lifecycle from dispatch to merge. (clankwork)

Expert in designing APIs, CLIs, and documentation systems for AI agent consumers. Use when building or reviewing interfaces intended to be used by LLM agents rather than humans — APIs, CLIs, SDKs, skills/docs, synchronization strategies, and concurrency models. Use when the user says "design an agent API", "make my API agent-friendly", "build a CLI for agents", "design skills/docs for agents", "agent interface review", or "how should agents use my product".