What is hermes-lab AGENTS.md?

Repository instructions for Hermes Lab, a framework for running and tracking experiments. The repository contains code and templates, while a separate data root stores experiment state, runs, and metrics.

Use it when entering Hermes Lab, locating its data root, following repository-ingress rules, or managing experiments through labctl.py.

It prevents agents from confusing project files with experiment data and defines the required reading and control commands.

How amenti-labs/hermes-lab configures Codex and OpenCode

Reuse

Borrowing it

Nothing to install: this file belongs to amenti-labs/hermes-lab. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/amenti-labs/hermes-lab/main/AGENTS.md

Clone the repo

git clone --depth 1 https://github.com/amenti-labs/hermes-lab

Made for: Codex, OpenCode.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/amenti-labs/hermes-lab/agents-md/github.svg)](https://agentmods.dev/instructions/amenti-labs/hermes-lab/agents-md)

Your own site

<a href="https://agentmods.dev/instructions/amenti-labs/hermes-lab/agents-md"><img src="https://agentmods.dev/badge/instructions/amenti-labs/hermes-lab/agents-md/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/instructions/amenti-labs/hermes-lab/agents-md"><img src="https://agentmods.dev/badge/instructions/amenti-labs/hermes-lab/agents-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 510 This file is loaded in full into every session.

When invoked 510 The same file — it is already loaded in full.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

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 5.1	$0.00510	$0.00510
Opus 5.5	$0.00204	$0.00204
Sonnet 5.5	$0.00102	$0.00102
Haiku 4.5	$0.00051	$0.00051

Measured 2mo ago against content hash dc92d29fca6c, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

hermes-lab 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 2mo 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.

AGENTS.md · 58 lines

How it starts

The opening of the file, as written. The whole thing — 58 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Hermes Lab agent instructions

Root-folder harness contract for AI coding and research agents

What this repo is

A generic autoresearch lab scaffold. The repo holds code and templates. The data root holds canonical experiment state.

python3 scripts/labctl.py is the control surface.
LAB_MANIFEST.json is the machine-readable command and contract reference.
README.md has the full user-facing docs (templates, search strategies, examples).

Data root vs workspace

Data root ($HERMES_LAB_DATA_ROOT, default ./lab-data): Lab state -- experiments, runs, metrics, dispatch queue.
Workspace (workspace_root in SPEC): Your experiment code. Can be any path on disk.

Repo ingress

Read these in order when pointed at the repo root:

README.md
LAB_MANIFEST.json
docs/AUTORESEARCH-COMPATIBILITY.md
docs/REFERENCE-EXECUTOR.md
docs/MULTI-FIDELITY.md
docs/LOCAL-AGENT-MUTATION.md
docs/DISPATCH.md

If a lab data root already exists, then read:

README-FIRST.md
LAB-STATUS.md
LAB-INDEX.json
PROGRAM.md
experiments/<id>/RUNBOOK.md
experiments/<id>/SUMMARY.md
experiments/<id>/NEXT.md

Write rules

Treat the data root as canonical state.
Only write inside a claimed run bundle when executing an experiment.
If working from dispatch/running/<dispatch-id>/, only write inside run/.
Do not hand-edit derived files.
RESULT.md is the write-ahead artifact.
Use the runbook and spec to determine mutable vs read-only surfaces.

Dispatch protocol

For agent-managed execution, use the dispatch protocol instead of writing directly into experiments/<id>/runs/. The queue package includes dispatch.json, an input/ snapshot, and a writable run/ folder that is ingested into canonical history only after completion.

If mutation_command is omitted and agent_provider is set, the lab synthesizes a local mutation worker through scripts/local_agent_mutation.py.

See docs/DISPATCH.md for the full protocol.

Read the full file on GitHub · 58 lines

Changes

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.

2mo ago First seen · 58 lines · 510 tokens per session scan A dc92d29fca6c

Subscribe to this mod's changes

hermes-lab AGENTS.md is an instructions file published in the GitHub repository amenti-labs/hermes-lab (5 stars, last pushed 7mo ago), licensed MIT. It adds 510 tokens to every session, about $0.0020 per session on Opus 5.5. 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.

Scientific

hermes-lab AGENTS.md next to 6 instructions from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
hermes-lab AGENTS.mdamenti-labs/hermes-lab	Instructions file	5	7mo ago	MIT	A	510	Codex, OpenCode
vscode buildNext.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	9,719	GitHub Copilot
next.js AGENTS.mdvercel/next.js	Instructions file	143k	yesterday	MIT	A	7,461	Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	5,649	GitHub Copilot
langchain AGENTS.mdlangchain-ai/langchain	Instructions file	149k	today	MIT	A	4,680	Codex, OpenCode
vscode copilot-instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	3,093	GitHub Copilot
vscode ai-customization.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	1,907	GitHub Copilot

hermes-lab: Instructions file for Codex