Code Review & Agent Automation Skills

Failure-mode-driven reliability engineering — enumerate how it breaks, give every mode a verdict, prove every handler with a reproducing test. Use when the user says "flaky", "harden this before launch", "what if this fails", "error handling", "timeout", "retry", "idempotency", or "race condition"; when an incident or postmortem lands; when a job crashes halfway and leaves partial state; when a webhook or queue consumer double-fires, sees double delivery, or charges a customer twice; when crash recovery is undefined; or when code is about to ship with only happy-path tests.

Launch, positioning, and promotion — making finished work land with the audience that should care. Use when the user says "launch", "announce", "release notes", "changelog", "blog post", "tweet thread", "X thread", "Show HN", "Hacker News post", "how do I promote this", "who would care about this", or "positioning"; when a README needs a pitch instead of a manual; when something is built and working but nobody outside the repo knows it exists; or when launch copy already exists but reads like a feature list.

Skill forging — distilling a team's recurring, hard-won practice into a new installable SKILL.md that meets the nen contract. Use when the user says "write a skill", "create a skill for", "turn this workflow into a skill", "codify our process", or "SKILL.md"; when someone says "we keep solving this the same way" or "the team keeps re-teaching this in review"; after a retro where the same lesson surfaces for the second or third time; or when an agent keeps making a mistake that a written discipline would have prevented.

Orchestration engineering for multi-agent and deep-research work — decomposition, routing, prompt design, and verified delegation. Use when the user says "plan this", "break this down", "delegate", "orchestrate", "subagent", "task distribution", "parallel workstreams", "split this across agents", "hand this to three agents", "write a prompt for", or "research X thoroughly"; when work must be divided into workstreams that different agents or people can own; when they ask which model or agent should handle something; when work is multi-step and ambiguous enough that one agent charging ahead would guess wrong; when several independent investigations could run in parallel; or when a subagent has reported "done" and someone is about to believe it.

Reverse engineering and mechanism-comparison research — when the docs run out, read the artifact itself and prove how the thing actually works. Use when the user asks "how does actually do it", says "reverse engineer this", "undocumented behavior", "the docs don't match what happens", or "the docs never mention it"; when a dependency, vendor SDK, legacy system, or external API behaves in ways nobody on the team can explain; when the ask is to compare implementations of the same mechanism across libraries or competitors; when someone says "we've tried everything" about a black-box system; or when the only source of truth is minified bundles, compiled binaries, or network traffic.

Behavior-preserving refactoring under characterization tests, plus severity-rated code review. Use when the user says "refactor", "clean up", "tech debt", "code review", "extract", "rename", "restructure", "this file is a mess", or "impossible to modify"; when every small change breaks something unrelated; when cleanup has to land before the next feature has anywhere to go; when code is called hard to change, scary to touch, or too tangled to test; when requirements shifted and the current shape fights the new feature; when a PR needs review for structure and maintainability; or when every small change keeps ballooning because everything touches everything.