Post-implementation audit of a master plan. Systematic dogfooding with uv run, /tmp copies, parallel stress testing, multi-round discovery, issue cataloging, and fix master plan generation. Use after implementing a master plan, when dogfooding, stress-testing, or ruthlessly testing implemented features.
CLI command patterns, global flags, project bootstrap, and workspace layout for the autopilot CLI. Use when adding commands, subcommands, modifying global flags, creating new workspaces, diagnosing workspace issues, or understanding the project overlay structure.
Deterministic, programmatic techniques for restructuring Python codebases -- splitting monoliths into packages, rewriting imports via AST, updating patch targets, and validating with quality gates. Use when reorganizing files, splitting large modules, moving code between packages, refactoring import paths, or restructuring directory layout.
Meta-planning skill for design exploration, POC verification, and iterative design refinement. Produces a meta plan document that feeds into create-master-plan (for multi-plan systems) or create-plan (for individual plans). Use when you want to think through a design before generating plan files, when POC validation is needed, or when the user asks to plan out a feature without generating plan files yet.
Feature planning guide for AutoPilot covering architecture boundaries, config layering, naming, testing, and subplan structure. Use when planning a new feature, designing a new module or command, or writing an implementation plan.
Comprehensive guide for authoring new Agent Skills in this repo following the agentskills.io specification. Use when creating a new skill, updating an existing skill definition, or reviewing skill quality.
Pre-implementation verification of master plans and standalone plans. Multi-pass auditing (coherence, ambiguity, DRY, simulation, readiness), fix plan generation, and readiness gates. Use when verifying a plan is implementation-ready, running audits, or checking plan coherence.