Borrowing it
Nothing to install: this file belongs to arcee-ai/nac. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/arcee-ai/nac/dev/AGENTS.mdgit clone --depth 1 https://github.com/arcee-ai/nacWrote 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.
[](https://agentmods.dev/instructions/arcee-ai/nac/agents-md)<a href="https://agentmods.dev/instructions/arcee-ai/nac/agents-md"><img src="https://agentmods.dev/badge/instructions/arcee-ai/nac/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.
<a href="https://agentmods.dev/instructions/arcee-ai/nac/agents-md"><img src="https://agentmods.dev/badge/instructions/arcee-ai/nac/agents-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.02185 | $0.02185 |
| Opus 5.5 | $0.00874 | $0.00874 |
| Sonnet 5.5 | $0.00437 | $0.00437 |
| Haiku 4.5 | $0.00218 | $0.00218 |
Grade A, and why
nac 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 4d 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 — 190 lines — stays where its author put it; the contents beside it link to each section on GitHub.
NAC repository guide
NAC is a Rust coding-agent harness with a React web client. It supports three immutable session behaviors: the established orchestrator/worker topology, a persistent direct coding agent with traditional child sessions, and a direct agent that can launch separate managed orchestrator sessions. Managed-host deployment is additive and opt-in.
Read the nearest nested AGENTS.md before changing a subsystem. This root file
defines repository-wide constraints; nested guides add ownership-specific
rules. Durable architectural decisions live in docs/architecture/. Historical
work ledgers and local notebooks are evidence, not a substitute for current code
and tests.
Dependency direction
Dependencies point inward:
React clients
-> HTTP/OpenAPI/MCP delivery (nac-server)
-> focused application services and composition ports
-> nac-core durable harness nac-managed product context
-> shared contracts/infrastructure adapters <-
- Domain and application contracts must not depend on Axum, React, provider wire types, or deployment wiring.
nac-manageddoes not depend onnac-serveror the agent harness. The server composes its ports with core/project services.- Provider, filesystem, database, Git/process, sandbox, and HTTP clients are adapters around inward contracts; do not move transport types into domain APIs.
- Model-visible tool exposure, authorization, hard safety policy, and the already-selected execution backend are separate decisions. Approval never changes or escapes the backend.
- Prefer private modules and narrow exports. Add a crate only for a real dependency boundary, not to increase crate count.
See dependency boundaries and generated API contract.
Where changes belong
| Change | Owner | Start here |
|---|---|---|
| Model loop, prompts, compaction | nac-core::agent |
crates/nac-core/src/agent/ |
| Session admission, recovery, cancellation, settlement | nac-core::session_service |
crates/nac-core/src/session_service/AGENTS.md |
| Durable schema, transcript, relationships, inbox/goals | nac-core::store and sessions |
crates/nac-core/src/store/AGENTS.md |
| Permission evaluation and approval | nac-core::permissions |
crates/nac-core/src/permissions/AGENTS.md |
| Native tools, capability composition, invocation | nac-core::tools |
crates/nac-core/src/tools/AGENTS.md |
| Model/runtime/backend construction | nac-core::runtime |
crates/nac-core/src/runtime/ |
| Managed secrets, GitHub, clone workflow, readiness | nac-managed |
crates/nac-managed/AGENTS.md |
| Product use cases and HTTP/OpenAPI/MCP delivery | nac-server |
crates/nac-server/AGENTS.md |
| React features, queries, presentation | web client | crates/nac-server/web/AGENTS.md |
| Managed image/runtime contract | managed container | docker/managed/AGENTS.md |
| Shared command environment and credentials | small inward crates | crates/nac-contracts/, crates/nac-credential-store/ |
| Descendant-aware process supervision | nac-process |
crates/nac-process/AGENTS.md |
| Checked-in model catalog generation | nac-catalog-gen |
crates/nac-catalog-gen/AGENTS.md |
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.
- 4d ago Changed · +4 lines · +58 tokens per session bb641b3fa851
- 10d ago Changed · +3 lines · +55 tokens per session 4eaef9faaa1f
- 24d ago Changed · +5 lines · +67 tokens per session b49baac421c9
- 1mo ago Changed · +2 lines · +35 tokens per session 9f3d51e03e22
- 1mo ago First seen · 176 lines · 1,970 tokens per session scan A 871694a4777d
nac AGENTS.md is an instructions file published in the GitHub repository arcee-ai/nac (296 stars, last pushed yesterday), licensed Apache-2.0. It adds 2,185 tokens to every session, about $0.0087 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 2026-09-12.
nac AGENTS.md compared with instructions from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| nac AGENTS.mdarcee-ai/nac | Instructions file | 296 | yesterday | Apache-2.0 | A | 2,185 | Codex, OpenCode |
| vscode buildNext.instructions.mdmicrosoft/vscode | Instructions file | 195k | yesterday | MIT | A | 9,719 | GitHub Copilot |
| next.js AGENTS.mdvercel/next.js | Instructions file | 143k | yesterday | MIT | A | 7,461 | Codex, OpenCode |
| langchain AGENTS.mdlangchain-ai/langchain | Instructions file | 149k | yesterday | MIT | A | 4,680 | Codex, OpenCode |
| vscode oss-third-party-notices.instructions.mdmicrosoft/vscode | Instructions file | 195k | yesterday | MIT | A | 5,516 | GitHub Copilot |
| vscode copilot-instructions.mdmicrosoft/vscode | Instructions file | 195k | yesterday | MIT | A | 3,093 | GitHub Copilot |
| deepseek-harness AGENTS.mddeepseek-ai/deepseek-harness | Instructions file | 274k | 2d ago | MIT | A | 4,313 | Codex, OpenCode |