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/frangelbarrera/agentic-harness/agents-mdgit clone --depth 1 https://github.com/frangelbarrera/agentic-harnessWrote 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/frangelbarrera/agentic-harness/agents-md)<a href="https://agentmods.dev/instructions/frangelbarrera/agentic-harness/agents-md"><img src="https://agentmods.dev/badge/instructions/frangelbarrera/agentic-harness/agents-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 | $0.01211 | $0.01211 |
| Opus 5 | $0.00606 | $0.00606 |
| Sonnet 5 | $0.00242 | $0.00242 |
| Haiku 4.5 | $0.00121 | $0.00121 |
Grade A, and why
agentic-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 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 — 94 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — System prompt for AI coding agents contributing to Agentic Harness
You are contributing to Agentic Harness, an open-source agent harness for Python.
Project Context
Agentic Harness is NOT a framework. It is a harness — the control layer that lets
developers orchestrate AI agents without surrendering their prompts, context,
model choice, or budget. Read MANIFESTO.md for the full philosophy.
Architecture in 30 seconds
- Thread: append-only event log (mutates in place for O(1) performance). State = reduce(events).
- Specialist: pre-built role-based agent (12 total: planner, coder, reviewer, tester, debugger, researcher, security-auditor, devops-engineer, data-scientist, product-manager, market-analyst, cost-estimator).
- Playbook Library: 13 domain templates (mobile_app, osint, financial_analysis, ...) + a TaskRouter that classifies natural-language requests into a domain without an LLM call.
- Playbook: YAML manual compiled to a DAG of steps. Each step invokes a specialist or tool. Steps can declare a
review:config for actor-critic iterative refinement. - Executor: walks the DAG, applies middleware (cost guard + verification + token optimizer), runs review loops when configured.
- MCP server: exposes playbooks as tools for Claude Desktop, Cursor, Cline, Zed.
Non-negotiable rules (from MANIFESTO.md)
- No vendor-only features as first-class APIs. If it only exists in OpenAI or only in Anthropic, it is a leak, not a feature.
- No classes named
Runnable,Chain,Workflow, orAgent. Composition = functions. - Token counter is on by default. If you don't know what you spent, you didn't ship.
- No hosted version. Ever. If you add code that requires Agentic Harness Cloud, you're breaking the manifesto.
- No hidden prompts. Every prompt sent to an LLM is a file on disk.
- No magic. If a line does something you can't explain, it's a bug.
- No API keys in code. Read from env. Never log. Never store.
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 · 94 lines · 1,211 tokens per session scan A 2857d932f649
agentic-harness AGENTS.md is an instructions file published in the GitHub repository frangelbarrera/agentic-harness (7 stars, last pushed 4d ago), licensed Apache-2.0. It adds 1,211 tokens to every session, about $0.0061 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
kungfu copilot-instructions.md
Copilot instructions for kungfu-systems/kungfu: Read and follow AGENTS.md before proposing or running changes. It is the shared agent router for this repository, including the canonical Shifu development and build entrypoint.
GPT-RAG release.instructions.md
Instructions for Azure/GPT-RAG, a project described as: Enterprise-grade accelerator for agentic RAG on Azure. Built on Microsoft Foundry with Foundry IQ as the default retrieval backend, Microsoft Agent Framework orchestration, Zero-Trust architecture and IaC.
autopus-adk GEMINI.md
Gemini CLI instructions for Insajin/autopus-adk, covering autopus-adk harness, 스킬 디렉터리, core guidelines, subagent delegation and file size limit.
aeon CLAUDE.md
Instructions for aeonfun/aeon, covering aeon, how aeon works, strategy, voice and soul file hierarchy (read in this order).
cognirepo CLAUDE.md
Claude Code instructions for ashlesh-t/cognirepo, covering claude.md, key rules, session start sequence (run in this order), behavioral confirmation rule and personas (cognirepo-402, cognirepo-403).
nanobot CLAUDE.md
Claude Code instructions for HKUDS/nanobot, a project described as: Ultra-lightweight, open-source, self-hosted personal AI agent framework in Python with WebUI, tools, memory, MCP, multi-agent workflows, automation, and chat apps.