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/ghosteken/agent-harness/agents-mdgit clone --depth 1 https://github.com/Ghosteken/agent-harnessWhat 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 | $0.03297 | $0.03297 |
| Opus 5 | $0.01648 | $0.01648 |
| Sonnet 5 | $0.00659 | $0.00659 |
| Haiku 4.5 | $0.00330 | $0.00330 |
Grade A, and why
agent-harness AGENTS.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 — 297 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
This file provides guidance to AI coding agents (Claude Code, Cursor, Copilot, Antigravity, etc.) when working with code in this repository.
Repository Overview
agent-harness is a production-grade collection of engineering skills and specialist agent personas for AI coding assistants (Claude Code, Cursor, Copilot, Gemini CLI, and others). Skills are step-by-step workflows that encode what senior engineers do consistently. Agent personas are specialist roles that apply those skills from a specific engineering perspective.
Agent Personas
twelve specialist personas are available in agents/. Each persona adopts a single engineering role and operates within its curated skill scope.
| Persona | Role | Trigger |
|---|---|---|
code-reviewer |
Senior Staff Engineer | Review any code change before merge |
security-auditor |
Security Engineer | Security audit, vulnerability detection |
test-engineer |
QA Engineer | Test strategy and coverage analysis |
web-performance-auditor |
Web Performance Engineer | Core Web Vitals and loading performance |
api-developer |
Senior API Engineer | API design, contract review, integration |
db-architect |
Senior Database Architect | Schema design, queries, migrations |
senior-devops |
Senior DevOps Engineer | Pipelines, IaC, containers, incidents |
senior-frontend-engineer |
Senior Frontend Engineer | UI components, accessibility, bundle size |
senior-cloud-architect |
Senior Cloud Architect | Cloud design, IAM, DR, cost |
senior-backend-engineer |
Senior Backend Engineer | Services, async, queues, caching |
ai-ml-engineer |
AI/ML Engineer | Prompts, RAG, LLM evaluation, AI safety |
senior-qa-engineer |
Senior QA Engineer | Feature verification, fix validation, regression checks |
Architecture: Command → invokes Agent persona → persona invokes Skill(s).
Each persona's Composition block (in its .md file) lists its curated skill scope — the 20 skills it may invoke. When an agent is active, invoke a skill from its scope by name: Use skill <skill-name> to.... Skills outside an agent's scope can be invoked manually or via a slash command at any time. The full scope registry is at scripts/agent-scopes.json.
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 · 297 lines · 3,297 tokens per session scan A 9857f14733af
agent-harness AGENTS.md is an instructions file published in the GitHub repository Ghosteken/agent-harness (2 stars, last pushed 17d ago), licensed MIT. It adds 3,297 tokens to every session, about $0.0165 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
gnap CLAUDE.md
Instructions for farol-team/gnap, covering claude.md, what this repo is, architecture, key files and core protocol concepts.
helmor AGENTS.md
Instructions for dohooo/helmor, covering agents.md, what is helmor, commands, architecture and three-process model.
zeroshot AGENTS.md
AGENTS.md instructions for the-open-engine/zeroshot, covering zeroshot: multi-agent coordination engine, critical rules, where to look, cluster worker contracts and cli quick reference.
agents CLAUDE.md
Instructions for eloylp/agents, covering claude.md, project overview, directory structure, config model and build & run.
AvaKit CLAUDE.md
Instructions for mericcintosun/AvaKit, covering claude.md — avakit, what this is, binding rules (do not violate), tech decisions (see docs/04-adr.md) and guardrails.
agiwo AGENTS.md
Instructions for xhwSkhizein/agiwo, covering agents.md, how to use this file, truth source order (adr 0048 / 0049), logging and config hot reload.