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/keesan12/martin-loop/agents-mdgit clone --depth 1 https://github.com/Keesan12/martin-loopWhat 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 | $0.01999 | $0.01999 |
| Opus 5 | $0.01000 | $0.01000 |
| Sonnet 5 | $0.00400 | $0.00400 |
| Haiku 4.5 | $0.00200 | $0.00200 |
Grade A, and why
martin-loop 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 yesterday.
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 — 252 lines — stays where its author put it; the contents beside it link to each section on GitHub.
MartinLoop Agent Operating Rules
These rules are mandatory for every AI agent, automation, contributor, and coding session working on MartinLoop.
Repository authority
is the sole development authority.
is a public distribution repository. It is never the primary development repository.
Mandatory internal-first sequence
Every code change must follow this sequence:
- Start from the latest .
- Create a private feature branch.
- Implement and test the change privately.
- Commit and push the private branch.
- Open a private PR into .
- Pass all required internal checks.
- Fix all regressions.
- Prove the real MartinLoop workflow works.
- Merge the private PR only after explicit approval.
- Record the private merge commit SHA.
- Verify fresh is clean and healthy.
- Create a clean public-staging branch from the latest public .
- Transfer only the reviewed public-safe diff from the private merge.
- Run public tests and public-surface guards.
- Push the public-staging branch.
- Open a public PR.
- Merge publicly only after hosted checks pass and explicit approval is given.
Prohibited actions
Agents must never:
- develop directly in ;
- use public as the implementation authority;
- push unmerged private feature commits to the public repository;
- create public staging before the private PR is merged;
- resolve implementation conflicts only in a public-staging branch;
- bypass or weaken the public-write hook;
- ask the user to bypass the hook for unfinished work;
- push directly to public ;
- merge a stale public branch wholesale;
- declare public readiness while private is broken;
- promote code without a recorded private merge SHA;
- publish npm, tags, native assets, or GitHub Releases before public merge approval.
Required internal health proof
Before public staging, a fresh worktree of private must prove all packages build and test clean, portability and copy-scan pass, git diff --check passes, no conflict markers exist, martin doctor works, preflight and governed run succeed, receipt is created and readable, verification succeeds, status survives restart, and separate workspaces remain isolated.
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.
- yesterday First seen · 252 lines · 1,999 tokens per session scan A 81df453c5098
martin-loop AGENTS.md is an instructions file published in the GitHub repository Keesan12/martin-loop (47 stars, last pushed yesterday), licensed Apache-2.0. It adds 1,999 tokens to every session, about $0.0100 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
ai-sdlc CLAUDE.md
Claude Code instructions for ai-sdlc-framework/ai-sdlc, covering ai-sdlc project instructions, git flow, ci marker hygiene, branches & commits and prs.
leashd CLAUDE.md
Instructions for vmehera123/leashd, covering claude.md, commands, install dependencies, run tests (single file / specific test / all) and run tests with coverage.
synkb CLAUDE.md
Claude Code instructions for Syntax753/synkb, covering claude.md, read these first, rules that are easy to get wrong, shell and testing.
regulator.ai AGENTS.md
AGENTS.md instructions for risk-ai/regulator.ai, covering agents.md - your workspace, first run, session startup, memory and 🧠 memory.md - your long-term memory.
VaultysClaw agent-controller.instructions.md
Use when working on the agent-controller package: tools, skills, agent runtime, WebSocket client, local SQLite DB, LLM integration, or CLI modes.
open-ace CLAUDE.md
Instructions for open-ace/open-ace, covering claude.md, where process documents go, pushing, test placement and ci semantics and schema snapshots (schema-sync ci).