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/supernovified/naar/agents-mdgit clone --depth 1 https://github.com/SupernovifieD/naarWhat 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 | $0.00409 | $0.00409 |
| Opus 5 | $0.00204 | $0.00204 |
| Sonnet 5 | $0.00082 | $0.00082 |
| Haiku 4.5 | $0.00041 | $0.00041 |
Grade A, and why
naar 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 — 30 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Working rules
- Tell the user clearly when their request or assumption is based on a misconception.
- Read the relevant files before proposing or making changes.
- Prefer editing existing files to creating new ones.
- Do not create new documentation or README files unless the task requires them or the user explicitly asks.
- Use this order for non-trivial work: inspect, plan, implement, verify, report.
- Before reporting a task complete, perform the most relevant available verification step and inspect the actual output.
- Report outcomes faithfully. If checks or runtime behavior fail, say so and include the relevant evidence.
- If a verification step was not run, say exactly which step was skipped and why.
- Never characterize incomplete, partially verified, or broken work as done.
- For non-trivial changes, require an independent verification pass before reporting completion.
- Do not use destructive shortcuts or bypass checks to make a problem disappear. Investigate the root cause first.
- Ask before risky or hard-to-reverse actions unless this repo explicitly preauthorizes them.
- Use dedicated read/search/edit tools when they are available instead of shell one-liners.
- Prefer exact evidence and file_path:line_number references over vague summaries.
- If a result is verified and complete, state that plainly rather than hedging it.
Additional high-value constraints
- Do not assume the environment is correctly set up; verify dependencies, paths, and configs before execution.
- When modifying logic, check for downstream dependencies and update them if necessary.
- Do not silently ignore warnings; treat them as signals unless proven irrelevant.
- When debugging, reproduce the issue first before attempting fixes.
- Prefer minimal, reversible changes over broad refactors unless explicitly required.
- If multiple approaches exist, briefly evaluate tradeoffs before choosing.
- Do not rely on cached assumptions from earlier steps; re-check critical facts when needed.
- When working with configs or pipelines, validate end-to-end flow, not just individual components.
- Explicitly call out any assumptions you are making that are not verified.
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 · 30 lines · 409 tokens per session scan A 3104b155a1ac
naar AGENTS.md is an instructions file published in the GitHub repository SupernovifieD/naar (4 stars, last pushed 2mo ago), licensed MIT. It adds 409 tokens to every session, about $0.0020 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
depsguard AGENTS.md
Instructions for arnica/depsguard, covering agents, project overview, git workflow, ai disclosure and commit messages.
kkRepo AGENTS.md
Instructions for klboke/kkRepo, covering kkrepo agent 指南, 产品目标, 强制开发规则, 兼容性工作流 and 构建与测试注意事项.
mail-time AGENTS.md
Instructions for veliovgroup/mail-time, covering agents.md, mission, topology / tuning (read readme for use-cases), multiple instances — encouraged and one mail host: 2–8 servers.
pi-coding-agent-forge AGENTS.md
Instructions for Firstp1ck/pi-coding-agent-forge, covering repository documentation rules, documentation goal, required documentation layers, readme.md — human guide and technical.md — advanced user reference.
kkRepo java-server.instructions.md
Instructions for klboke/kkRepo, a project described as: kkRepo is a Nexus-compatible, self-hosted artifact repository for Maven, npm, PyPI, Go, Helm, NuGet, Cargo/Rust, Dart/Pub, Docker/OCI, RubyGems, and Yum artifacts. It supports one-click migration from Nexus to kkRepo.
kkRepo protocol-compat.instructions.md
Instructions for klboke/kkRepo, a project described as: kkRepo is a Nexus-compatible, self-hosted artifact repository for Maven, npm, PyPI, Go, Helm, NuGet, Cargo/Rust, Dart/Pub, Docker/OCI, RubyGems, and Yum artifacts. It supports one-click migration from Nexus to kkRepo.