Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/jrenaldi79/harness-engineering/claude-mdgit clone --depth 1 https://github.com/jrenaldi79/harness-engineeringWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/jrenaldi79/harness-engineering/claude-md)<a href="https://agentmods.dev/instructions/jrenaldi79/harness-engineering/claude-md"><img src="https://agentmods.dev/badge/instructions/jrenaldi79/harness-engineering/claude-md.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.02791 | $0.02791 |
| Opus 5 | $0.01396 | $0.01396 |
| Sonnet 5 | $0.00558 | $0.00558 |
| Haiku 4.5 | $0.00279 | $0.00279 |
Grade A, and why
harness-engineering CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 213 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
This file provides guidance to Claude Code when working with code in this repository.
Project Overview
harness-engineering is a Claude Code plugin and reference for AI coding agent harnesses. It provides two skills (/readiness and /setup) that analyze and configure projects for agent-assisted development, plus a README mapping 20+ best practices from industry sources.
Core Features
/readiness: Scores a codebase across 8 pillars and 5 maturity levels, produces a saved report with delta tracking/setup: Scaffolds CLAUDE.md files, enforcement scripts, git hooks, linter configs, and agent settings via Socratic questioning- Reference guide: Maps best practices from OpenAI, Anthropic, Augment Code, Factory.ai, and practitioners to concrete implementation patterns
Essential Commands
Testing
node --experimental-vm-modules node_modules/.bin/jest tests/scripts/ # Unit tests for setup scripts
bash tests/evals/run-evals.sh # E2E readiness evals (default)
bash tests/evals/run-evals.sh --config setup-eval-config.json # E2E setup evals
bash tests/evals/test-marketplace-install.sh # Test plugin install flow
Validation
node skills/setup/scripts/lib/generate-docs.js --check # Verify auto-generated sections are current
node skills/setup/scripts/lib/validate-docs.js --full # Check for documentation drift
Setup
bash scripts/install-hooks.sh # Install git hooks (pre-commit + pre-push)
Architecture
skills/ ├── readiness/ │ └── SKILL.md └── setup/ ├── references/ │ ├── claude-md-guide.md │ ├── enforcement-scripts.md │ └── stack-node-typescript.md ├── scripts/ │ ├── hooks/ │ │ ├── pre-commit │ │ └── pre-push │ ├── lib/ │ │ ├── check-file-sizes.js # File size enforcement script for pre-commit hook. │ │ ├── check-secrets.js # Secret detection script for pre-commit hook. │ │ ├── check-test-colocation.js # Test colocation enforcement script for pre-commit hook. │ │ ├── generate-docs-helpers.js # Helper functions for generate-docs.js. │ │ ├── generate-docs.js # Auto-generate CLAUDE.md sections from source code. │ │ └── validate-docs.js # CLAUDE.md drift detection script. │ ├── generate-claude-md.js # Generate tailored CLAUDE.md files for a project from templates. │ ├── init-project.js # Project scaffolding script for Node/TypeScript projects. │ └── install-enforcement.js # Copies enforcement tooling into a target project. ├── templates/ │ ├── rules/ │ │ ├── code-quality.md │ │ ├── react.md │ │ ├── tdd.md │ │ ├── testing.md │ │ └── typescript.md │ ├── eslint-base.js │ ├── gitignore-template │ ├── global-claude.md │ ├── lint-staged.config.js │ ├── project-claude.md │ └── settings.json └── SKILL.md scripts/ ├── hooks/ │ ├── pre-commit │ └── pre-push ├── install-hooks.sh ├── README.md ├── release.sh └── repo-generate-docs.js # Repo-level CLAUDE.md auto-generator. tests/ ├── evals/ │ ├── eval-config.json │ ├── grader.js # Readiness Skill Grader │ ├── hook-commit-validator.js # Hook-driven commit validation for setup eval grader. │ ├── README.md │ ├── run-evals.sh │ ├── setup-eval-config.json │ ├── setup-grader.js # Setup Skill Grader — validates /setup output against setup-eval-config.json. │ ├── setup-readiness-eval-config.json │ ├── setup-readiness-grader.js # Setup-then-Readiness Grader — validates that /setup produces a project │ └── test-marketplace-install.sh └── scripts/ ├── check-file-sizes.test.js # Tests for skills/setup/scripts/lib/check-file-sizes.js ├── check-secrets.test.js # Tests for skills/setup/scripts/lib/check-secrets.js ├── check-test-colocation.test.js # Tests for skills/setup/scripts/lib/check-test-colocation.js ├── detect-source-dirs.test.js # Tests for detectSourceDirs and buildModuleIndex adaptive scanning. ├── doc-drift-detection.test.js # Tests for doc drift detection — verifies validate-docs.js catches ├── doc-indexing-roundtrip.test.js # Tests for post-install doc indexing round-trip — verifies that ├── enforcement-roundtrip.test.js # Tests for post-install enforcement script round-trip — verifies that scripts ├── generate-claude-md.test.js # Tests for skills/setup/scripts/generate-claude-md.js ├── generate-docs-helpers.test.js # Tests for generate-docs-helpers.js: directory trees, module indexes, ├── generate-docs.test.js # Tests for generate-docs.js marker operations: replaceMarkers, ├── hook-integration.test.js # Tests for git commit hook integration — verifies that git commit triggers ├── incremental-doc-indexing.test.js # Tests for incremental doc indexing — verifies that adding a new source file ├── init-project.test.js # Tests for skills/setup/scripts/init-project.js ├── install-enforcement.test.js # Tests for skills/setup/scripts/install-enforcement.js ├── marketplace-schema.test.js # Tests for .claude-plugin/marketplace.json schema validity. ├── plugin-cache-recursion.test.js # Tests for plugin cache recursion bug (ENAMETOOLONG). ├── README.md ├── release.test.js # Tests for scripts/release.sh — validates version bumping, changelog ├── repo-generate-docs.test.js # Tests for scripts/repo-generate-docs.js — the repo-level CLAUDE.md └── validate-docs.test.js # Tests for skills/setup/scripts/lib/validate-docs.js
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 4d ago First seen · 213 lines · 2,791 tokens per session scan A ae50f040ffd1
harness-engineering CLAUDE.md is an instructions file published in the GitHub repository jrenaldi79/harness-engineering (66 stars, last pushed 5mo ago), licensed MIT. It adds 2,791 tokens to every session, about $0.0140 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
vibe-27 AGENTS.md
AGENTS.md instructions for fatalex27/vibe-27, covering vibe-27 behavioral guidelines, 1. editor, not typist, 2. intention over implementation, 3. trust and verify and 4. think before coding.
lynxprompt-vscode CLAUDE.md
Claude Code instructions for GeiserX/lynxprompt-vscode, covering claude.md — lynxprompt vs code, tech stack, development, architecture and key commands.
nv-context CLAUDE.md
Instructions for johnnichev/nv-context, covering claude-specific: nv-context, session management, compounding engineering and claude-specific.
nv-context GEMINI.md
Instructions for johnnichev/nv-context, covering nv:context — gemini cli support, skill and tool mapping.
nv-context AGENTS.md
Instructions for johnnichev/nv-context, covering nv-context, commands, install skill globally (recommended), install skill manually and test skill works (open any project, then run).
sourcebook CLAUDE.md
Claude Code instructions for maroondlabs/sourcebook, covering claude.md, commands, critical constraints, architecture and deprecated patterns.