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/lfaoro/ssm/agents-mdgit clone --depth 1 https://github.com/lfaoro/ssmWhat 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 | $0.01989 | $0.01989 |
| Opus 5 | $0.00994 | $0.00994 |
| Sonnet 5 | $0.00398 | $0.00398 |
| Haiku 4.5 | $0.00199 | $0.00199 |
Grade A, and why
ssm 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 — 129 lines — stays where its author put it; the contents beside it link to each section on GitHub.
SSM — Agent Working Instructions
This document exists to make AI coding agents (and humans) effective and safe when modifying the codebase. It prioritizes actionable rules, invariants, and verification steps over descriptive snapshots of the current state.
The Prime Directive
NEVER commit changes unless explicitly instructed.
Prepare the staging area (git add), show the plan and the full diff, and wait for the user to explicitly say "commit" or "go ahead". This rule overrides all other instructions. Do not ask the user whether they want to commit.
1. Verification — Run These Before Claiming Work Is Complete
An agent must produce clean results from the project's canonical checks before presenting changes as done.
Primary command (use this most often):
make check
make check runs (see Makefile):
gofmtlint(golangci-lint if present, elsego fmt + go vet)go-mod-tidy-checkmake test(go test -race -count=1 ./...)make build
Individual commands you will use constantly:
go build ./...
go vet ./...
go test -race -count=1 ./...
golangci-lint run ./...
govulncheck ./...
make build-static
When touching performance-sensitive code, also consider make bench (and the -cpu / -mem variants).
Rule: If you have not run the relevant verification commands and shown the output, you are not finished.
2. Non-Negotiable Rules
- Changelog discipline: Any commit that changes code, fixes bugs, adds features, or adjusts tests must also update
CHANGELOG.mdwith a concise entry under the appropriate keepachangelog.com section (Security, Fix, Add, Refactor, Test, Docs, etc.). - Injection prevention: The
--delimiter must appear before every hostname in all SSH, mosh, andsyscall.Execinvocations. This is non-negotiable. - SFTP connection flags: SFTP and certain remote operations deliberately use
BatchMode=yes,RequestTTY=no, andStrictHostKeyChecking=no. These are intentional (users are connecting to their own servers). See SECURITY.md. - Sensitive data handling:
IdentityFile,ProxyCommand,CertificateFile, and similar keys must remain filtered from any configuration display/viewport. - Parser thread-safety contract: See Architecture Contracts below. Do not regress the brief-publish-lock model.
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 · 129 lines · 1,989 tokens per session scan A 458b586dee76
ssm AGENTS.md is an instructions file published in the GitHub repository lfaoro/ssm (112 stars, last pushed 8d ago), licensed MIT. It adds 1,989 tokens to every session, about $0.0099 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
zot AGENTS.md
Instructions for patriceckhart/zot, covering working agreement for zot, product intent, starting a task, code ownership map and correctness contracts.
pvetui AGENTS.md
Instructions for devnullvoid/pvetui, covering agent instructions, initial setup, development workflow, quick reference and code quality standards.
waveloom AGENTS.md
Instructions for Menfre01/waveloom, covering waveloom, 项目概要, 编码规范, 代码审查 and 开发流程.
pvetui CLAUDE.md
Instructions for devnullvoid/pvetui: Canonical instructions live in AGENTS.md.
microNeo AGENTS.md
Instructions for sollawen/microNeo, covering 基本规则, debug and 核心概念.
tmuxwatch AGENTS.md
Instructions for steipete/tmuxwatch, covering agents.md, codex global instructions, general guardrails, intake & scoping and tooling & command wrappers.