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/harnessprotocol/harness-kit/claude-mdgit clone --depth 1 https://github.com/harnessprotocol/harness-kitWhat 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 | $0.01579 | $0.01579 |
| Opus 5 | $0.00790 | $0.00790 |
| Sonnet 5 | $0.00316 | $0.00316 |
| Haiku 4.5 | $0.00158 | $0.00158 |
Grade A, and why
harness-kit 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 2d 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 — 163 lines — stays where its author put it; the contents beside it link to each section on GitHub.
harness-kit
Claude Code configuration — skills, agents, and settings worth sharing, distributed as a plugin marketplace.
Repo structure
harness-kit/
├── .claude-plugin/
│ └── marketplace.json ← marketplace catalog
├── .github/
│ ├── schema/
│ │ └── plugin.schema.json ← Protocol JSON Schema for plugin manifest validation
│ └── workflows/
│ └── validate.yml ← CI: manifest parsing, version alignment, schema validation, builds, tests
├── plugins/ ← 17 plugins; see .claude-plugin/marketplace.json for full list
├── packages/ ← shared libraries
│ ├── core/ ← harness.yaml compile/parse/detect logic
│ ├── shared/ ← shared TypeScript types used across apps
│ ├── design-tokens/ ← single source of truth for the shared color identity; generates CSS vars for app + website
│ └── marketplace-data/ ← build-time generator: git → static marketplace JSON (runs the security scanner)
├── functions/ ← Cloudflare Pages Functions (server-side endpoints, e.g. /feedback proxy)
├── website/ ← documentation site + plugin marketplace (harnesskit.ai)
├── homebrew/ ← Homebrew tap formulae for CLI and desktop app
├── apps/ ← end-user applications
│ ├── desktop/ ← Tauri desktop app (React + Rust)
│ └── cli/ ← harness CLI
├── profiles/ ← pre-configured harness.yaml bundles for different roles
├── evals/ ← automated skill evaluation framework (Python + golden responses)
├── harness.yaml ← dogfooding config (plugins used to develop this repo)
├── install.sh ← script fallback for users without plugin marketplace
├── CLAUDE.md ← this file
└── README.md
Adding a new plugin
1. Create the plugin directory
plugins/<plugin-name>/
├── .claude-plugin/
│ └── plugin.json
├── agents/ ← optional, for specialist subagents
├── scripts/ ← optional, for bundled utilities
└── skills/
└── <skill-name>/
├── SKILL.md
└── README.md
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.
- 2d ago First seen · 163 lines · 1,579 tokens per session scan A e12345b5887b
harness-kit CLAUDE.md is an instructions file published in the GitHub repository harnessprotocol/harness-kit (10 stars, last pushed 3d ago), licensed Apache-2.0. It adds 1,579 tokens to every session, about $0.0079 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 2026-08-31.
Other instructions, from other repositories
raytsystem-public-os CLAUDE.md
Instructions for romarayt/raytsystem-public-os, covering compatibility pointer and documentation synchronization.
raytsystem-public-os AGENTS.md
Instructions for romarayt/raytsystem-public-os, covering raytsystem — agent routing, invariants, commands, skill routing and documentation.
obsidian-pkm-plugin CLAUDE.md
Instructions for AdrianV101/obsidian-pkm-plugin, covering pkm: 01-projects/obsidian-mcp, claude.md, project overview, commands and install dependencies.
gnosys AGENTS.md
Instructions for proticom/gnosys, covering agents.md, what is gnosys, quick reference, architecture and key rules.
brain-starter CLAUDE.md
Instructions for michaeljauk/brain-starter, covering brain repo - claude instructions, purpose, not set up yet?, file placement rules and directory guide.
skills AGENTS.md
Instructions for citedy/skills, covering agent instructions — citedy skills, repository, directory structure, adding a new skill and skill types.