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 |
First Move
- Inspect what already exists: instruction files, feature/state files, verification commands, docs, package manifests.
- Ask only for missing context that cannot be inferred safely: target agent, desired file name, tolerance for structure, and whether overwriting is allowed.
- Prefer a minimal harness first. Add memory, tool safety, multi-agent, or benchmark details only when the user's problem calls for them.
Common Tasks
Create a harness
Use the bundled script when working on a local repository:
node skills/harness-creator/scripts/create-harness.mjs --target /path/to/project
Options:
--agent-file CLAUDE.md for Claude-oriented projects.--package-manager npm|pnpm|yarn|bun when detection is wrong.--commands "cmd one,cmd two" for custom verification.--force only after confirming overwrites are acceptable.
Then explain what was created and how the user should replace placeholder feature entries.
Audit an existing harness
Run:
node skills/harness-creator/scripts/validate-harness.mjs --target /path/to/project
Report the five subsystem scores, the lowest-scoring area, and the first 2-3 changes that would improve reliability. Treat the lowest score as a candidate bottleneck; confirm with failures, logs, or task outcomes before claiming causality.
Produce a report
Use when the user wants a shareable assessment:
node skills/harness-creator/scripts/render-assessment-html.mjs --target /path/to/project
node skills/harness-creator/scripts/run-benchmark.mjs --target /path/to/project --html /path/to/report.html
Be clear that this is a structural benchmark. The benchmark first runs a self-check — it scaffolds a throwaway harness and validates it, proving the bundled scripts work end-to-end — then scores the target and eval coverage. Real effectiveness still needs before/after agent sessions on representative tasks.
When to Read References
Load only the reference needed for the user's problem:
- Memory across sessions: Memory Persistence
- Reusable workflows as skills: Skill Runtime
- Permissions, tools, concurrency: Tool Registry & Safety
- Context budget and progressive disclosure: Context Engineering
- Delegation and parallel agents: Multi-Agent Coordination
- Hooks, startup, long-running work: Lifecycle & Bootstrap
- Non-obvious failure modes: Gotchas
Design Rules
- Keep the root instruction file short: routing and invariants, not a full manual.
- Put project facts in project docs, not in the skill.
- Make verification commands explicit and runnable.
- Require evidence before marking a feature done.
- Use one active feature unless the harness has explicit multi-agent ownership boundaries.
- Prefer append/update state files over relying on chat history.
- Never hide destructive behavior in scripts; overwrites require explicit user approval.
Deliverable Checklist
For a usable minimal harness, leave the target project with:
If you cannot create files, provide exact file contents and commands instead.
1---2name: harness-creator3description: Build, audit, and improve harnesses that make AI coding agents reliable: AGENTS.md/CLAUDE.md instruction files, feature/state tracking, verification gates, scope boundaries, session handoff, memory persistence, context budgets, tool-permission safety, and multi-agent coordination. Use this whenever a coding agent is unreliable across sessions — forgets context, drifts out of scope, claims "done" before tests pass, or starts each session inconsistently — or when creating or assessing AGENTS.md, CLAUDE.md, feature_list.json, init.sh, progress.md, or session-handoff files. Reach for it even if the user never says the word "harness."4license: MIT5---67# Harness Creator89Use 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.1011Not for model selection, prompt tuning in isolation, chat UI design, or general app architecture.1213## Core Model1415Every useful coding-agent harness has five subsystems:1617| Subsystem | Minimal artifact | Purpose |18|---|---|---|19| Instructions | `AGENTS.md` or `CLAUDE.md` | Startup path, working rules, definition of done |20| State | `feature_list.json`, `progress.md` | Current feature, status, evidence, next step |21| Verification | `init.sh` or documented commands | Tests/checks the agent must run before claiming done |22| Scope | Feature dependencies and done criteria | Prevents overreach and half-finished work |23| Lifecycle | `session-handoff.md`, end-of-session routine | Makes the next session restartable |2425## First Move26271. Inspect what already exists: instruction files, feature/state files, verification commands, docs, package manifests.282. Ask only for missing context that cannot be inferred safely: target agent, desired file name, tolerance for structure, and whether overwriting is allowed.293. Prefer a minimal harness first. Add memory, tool safety, multi-agent, or benchmark details only when the user's problem calls for them.3031## Common Tasks3233### Create a harness3435Use the bundled script when working on a local repository:3637```bash38node skills/harness-creator/scripts/create-harness.mjs --target /path/to/project39```4041Options:4243- `--agent-file CLAUDE.md` for Claude-oriented projects.44- `--package-manager npm|pnpm|yarn|bun` when detection is wrong.45- `--commands "cmd one,cmd two"` for custom verification.46- `--force` only after confirming overwrites are acceptable.4748Then explain what was created and how the user should replace placeholder feature entries.4950### Audit an existing harness5152Run:5354```bash55node skills/harness-creator/scripts/validate-harness.mjs --target /path/to/project56```5758Report the five subsystem scores, the lowest-scoring area, and the first 2-3 changes that would improve reliability. Treat the lowest score as a candidate bottleneck; confirm with failures, logs, or task outcomes before claiming causality.5960### Produce a report6162Use when the user wants a shareable assessment:6364```bash65node skills/harness-creator/scripts/render-assessment-html.mjs --target /path/to/project66node skills/harness-creator/scripts/run-benchmark.mjs --target /path/to/project --html /path/to/report.html67```6869Be clear that this is a structural benchmark. The benchmark first runs a self-check — it scaffolds a throwaway harness and validates it, proving the bundled scripts work end-to-end — then scores the target and eval coverage. Real effectiveness still needs before/after agent sessions on representative tasks.7071## When to Read References7273Load only the reference needed for the user's problem:7475- Memory across sessions: [Memory Persistence](references/memory-persistence-pattern.md)76- Reusable workflows as skills: [Skill Runtime](references/skill-runtime-pattern.md)77- Permissions, tools, concurrency: [Tool Registry & Safety](references/tool-registry-pattern.md)78- Context budget and progressive disclosure: [Context Engineering](references/context-engineering-pattern.md)79- Delegation and parallel agents: [Multi-Agent Coordination](references/multi-agent-pattern.md)80- Hooks, startup, long-running work: [Lifecycle & Bootstrap](references/lifecycle-bootstrap-pattern.md)81- Non-obvious failure modes: [Gotchas](references/gotchas.md)8283## Design Rules8485- Keep the root instruction file short: routing and invariants, not a full manual.86- Put project facts in project docs, not in the skill.87- Make verification commands explicit and runnable.88- Require evidence before marking a feature done.89- Use one active feature unless the harness has explicit multi-agent ownership boundaries.90- Prefer append/update state files over relying on chat history.91- Never hide destructive behavior in scripts; overwrites require explicit user approval.9293## Deliverable Checklist9495For a usable minimal harness, leave the target project with:9697- [ ] `AGENTS.md` or `CLAUDE.md`98- [ ] `feature_list.json`99- [ ] `progress.md`100- [ ] `init.sh`101- [ ] Optional `session-handoff.md` for multi-session work102- [ ] Documented verification evidence or next action103104If you cannot create files, provide exact file contents and commands instead.
om-scogo