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/agent-sh/agentsys/agents-mdgit clone --depth 1 https://github.com/agent-sh/agentsysWhat 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 | $0.03115 | $0.03115 |
| Opus 5 | $0.01558 | $0.01558 |
| Sonnet 5 | $0.00623 | $0.00623 |
| Haiku 4.5 | $0.00312 | $0.00312 |
Grade A, and why
agentsys 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 — 278 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Project Memory: AgentSys
Cross-tool compatible: Works across all 5 platforms - Claude Code, OpenCode, Codex CLI, Cursor, and Kiro.
-
Production project - Real users depend on this. Test thoroughly and verify all edge cases before committing. WHY: Breaking changes affect all plugin users immediately.
-
Plugin for OTHER projects - Optimize for plugin users, not internal dev convenience. WHY: Every decision MUST improve the experience for developers using this in their repos.
-
Use CHANGELOG.md for completion tracking - MUST use CHANGELOG.md for release notes. NEVER create
*_AUDIT.md,*_SUMMARY.md,*_COMPLETION.mdfiles. WHY: Summary files clutter repos and add no value. Report completion verbally. -
Unless it is a very small change of max few lines, or an urgent hotfix, MUST create PRs for all changes - No direct pushes to main. WHY: PRs enable reviews, CI checks, and rollback if needed. Direct pushes are risky.
-
PR reviews - Wait 3 min for auto-reviewers, address ALL comments (Copilot, Claude, Gemini, Codex). WHY: Skipping comments leads to merged issues. Every comment must be addressed or explained.
-
Read checklists BEFORE multi-file changes - MUST read the relevant checklist before starting:
- Cross-platform work →
checklists/cross-platform-compatibility.md(MASTER REF) - Release →
checklists/release.md - New command →
checklists/new-command.md - New agent →
checklists/new-agent.md - New skill →
checklists/new-skill.md - New lib module →
checklists/new-lib-module.md - OpenCode plugin update →
checklists/update-opencode-plugin.md - Repo intel changes →
checklists/repo-intel.mdWHY: Multi-file changes have hidden dependencies. Checklists prevent missed updates.
- Cross-platform work →
-
Especially Before release, and when delivering ANY work - Check the relevant checklist for completion requirements:
- Identify which checklist applies to your work (see rule #6)
- Go through EVERY item in that checklist and apply it (e.g. run commands, which files to bump versions in, etc)
- Run the
/enhancecommand on new or modified commands, agents, skills, hooks or prompts - Verify cross-platform compatibility (OpenCode + Codex)
- Update
bin/cli.jsmappings if new command/agent added - Only mark complete after ALL checklist items are done WHY: Checklists exist because we kept missing things. They are the definition of "done".
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 · 278 lines · 3,115 tokens per session scan A 24a2c0b740bd
agentsys AGENTS.md is an instructions file published in the GitHub repository agent-sh/agentsys (980 stars, last pushed 5d ago), licensed MIT. It adds 3,115 tokens to every session, about $0.0156 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
Upsonic CLAUDE.md
Claude Code instructions for Upsonic/Upsonic, covering claude.md, project overview, ai operational guides, default pre-work consultation and keep documents/ai/explanation/ in sync with code.
intelligent-terminal agent-skills.instructions.md
Guidelines for creating high-quality Agent Skills for GitHub Copilot.
connectonion AGENTS.md
AGENTS.md instructions for openonion/connectonion, covering agents.md, project overview, architecture, core components and key design patterns.
ironcurtain CLAUDE.md
Instructions for provos/ironcurtain, covering claude.md, general workflow, git & worktrees, git workflow and platform considerations.
cyrus copilot-instructions.md
Copilot instructions for cyrusagents/cyrus: Note, there is a need to maintain the use of '--print' when running the claude exec commands because that is what makes it non-interactive.
AgentEval copilot-instructions.md
Instructions for AgentEvalHQ/AgentEval, covering agenteval - ai coding agent instructions, architecture overview, environment setup, optional: secondary models for comparison and build & test commands.