Harness Creator

If you've ever watched a coding agent lose track of what it was doing or forget its constraints halfway through a session, this is the fix. It helps you build the five things agents actually need: a startup file like AGENTS.md, a feature list with status, verification commands they must run, scope boundaries to prevent overreach, and session handoff notes so the next run picks up cleanly. Ships with scripts to create, audit, or benchmark a harness. The design is opinionated: keep instructions short, require evidence before marking work done, and make state explicit instead of buried in chat history. Comes with references on memory persistence, tool safety, and multi-agent coordination when you need them.

skills.sh installs: 2.5K
GitHub stars: 24.1K
last updated: Jul 14, 2026

View on GitHub

Install

$ npx -y skills add walkinglabs/learn-harness-engineering --skill harness-creator --agent claude-code

Installs into .claude/skills of the current project.

Files

Harness Creator

Use this skill to make a repository easier for coding agents to start, stay in scope, verify work, and resume across sessions. Keep the harness small enough that agents actually follow it.

Not for model selection, prompt tuning in isolation, chat UI design, or general app architecture.

Core Model

Every useful coding-agent harness has five subsystems:

Subsystem	Minimal artifact	Purpose
Instructions	`AGENTS.md` or `CLAUDE.md`	Startup path, working rules, definition of done
State	`feature_list.json`, `progress.md`	Current feature, status, evidence, next step
Verification	`init.sh` or documented commands	Tests/checks the agent must run before claiming done
Scope	Feature dependencies and done criteria	Prevents overreach and half-finished work
Lifecycle	`session-handoff.md`, end-of-session routine	Makes the next session restartable