harness-engineering

Mandatory routing/meta skill for the harness-engineering plugin. Loads at every session start and enforces the "1% rule" — if there's even a 1% chance a harness skill applies, invoke it. Routes between 19 skills (13 vendored from superpowers v5.1.0 + 6 harness-original).

Use when creating new skills, editing existing skills, or verifying skills work before deployment

Use when starting feature work that needs isolation from current workspace or before executing implementation plans - ensures an isolated workspace exists via native tools or git worktree fallback

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes

Use when executing implementation plans with independent tasks in the current session

Use when completing tasks, implementing major features, or before merging to verify work meets requirements

Use when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires technical rigor and verification, not performative agreement or blind implementation

Use when implementation is complete, all tests pass, and you need to decide how to integrate the work - guides completion of development work by presenting structured options for merge, PR, or cleanup

Use when you have a written implementation plan to execute in a separate session with review checkpoints

Use when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies

You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.

Use when implementing any feature or bugfix, before writing implementation code

Use when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always

Use when you have a spec or requirements for a multi-step task, before touching code

AI Agent Harness engineering initialization for new projects. Activate when users mention "new project", "project initialization", "set up Harness", "create CLAUDE.md", "set up Agent environment", "init harness", "start from scratch", "initialize AI coding environment", "establish Agent constraints", or "set up Claude Code project". Even if the user simply mentions wanting to "start a new project" or "configure an AI development environment", use this Skill, because every new project should begin by establishing a Harness.

Health check and optimization for existing project Harness setups. Activate when the user mentions "check Harness", "optimize CLAUDE.md", "Agent keeps making mistakes", "Harness health", "audit Harness", "evaluate AI coding environment", "harness audit", "check Agent config", "why won't the Agent follow instructions", "improve Agent effectiveness", "add Harness to existing project", or "legacy optimization". Also use this Skill when the user complains that Agent behavior deviates from expectations, the Agent repeatedly makes the same mistakes, or the project has been around for a while but lacks a systematic Harness framework — diagnose and improve.

Continuous iteration and evolution of the Harness framework. Activate when the user mentions "trim Harness", "remove unnecessary rules", "update Harness after new model release", "Harness maintenance", "garbage collection", "code entropy", "documentation drift", "architecture drift", "CLAUDE.md is too long", "harness evolve", "slim down Harness", "optimize token usage", or "clean up outdated rules". Also use this Skill when the user has updated the Claude model version, the project has gone through a long iteration cycle, or the Harness feels bloated and no longer efficient — use it to streamline and evolve the Harness framework.