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/arnica/depsguard/agents-mdgit clone --depth 1 https://github.com/arnica/depsguardWrote 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/arnica/depsguard/agents-md)<a href="https://agentmods.dev/instructions/arnica/depsguard/agents-md"><img src="https://agentmods.dev/badge/instructions/arnica/depsguard/agents-md.svg" alt="Measured on agentmods" 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 | $0.02669 | $0.02669 |
| Opus 5 | $0.01334 | $0.01334 |
| Sonnet 5 | $0.00534 | $0.00534 |
| Haiku 4.5 | $0.00267 | $0.00267 |
Grade A, and why
depsguard 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 — 218 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agents
Guidelines for AI agents working on this codebase.
Project overview
DepsGuard is a zero-dependency Rust CLI that scans package manager configs (npm, pnpm, bun, uv) for supply chain security best practices and offers interactive fixes. It targets Linux, macOS, and Windows.
Git workflow
- Never push directly to
mainwithout explicit permission from the user. Always use a feature branch and open a pull request. - Do not force-push or run destructive git operations unless explicitly asked.
AI disclosure
All agent involvement in this repository must be disclosed:
- Pull requests: every PR must state whether an AI agent or assistant produced any part of it (code, tests, docs, commit messages), naming the tool and model (e.g. "Claude Code — Claude Fable 5"). The PR template has a section for this; write "None" when no agent was used.
- Commits: agent-authored or agent-assisted commits must carry a
Co-Authored-Bytrailer naming the agent and model, e.g.Co-Authored-By: Claude Fable 5 <[email protected]>. - Comments and reviews: when an agent posts on a person's behalf, it must identify itself as an agent and name the model, e.g. "Authored by Claude (Fable 5) via Claude Code, on behalf of @user".
Commit messages
Use Conventional Commits (https://www.conventionalcommits.org/).
Format: <type>(<optional scope>): <description>
Types: feat, fix, refactor, test, docs, chore, ci, perf, style, build.
Examples:
feat(manager): add yarn berry supportfix(fix): preserve comments when writing .npmrctest: add integration tests for bun configdocs: update supported managers tablerefactor(term): simplify Windows console FFI
Rules:
- Use lowercase for the description; no trailing period.
- Keep the subject line under 72 characters.
- Use the body (separated by a blank line) to explain why, not what, when the change is non-trivial.
- Reference issue numbers in the footer when applicable (
Closes #42).
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 First seen · 218 lines · 2,669 tokens per session scan A 44d49e7cdfaf
depsguard AGENTS.md is an instructions file published in the GitHub repository arnica/depsguard (381 stars, last pushed 22d ago), licensed MIT. It adds 2,669 tokens to every session, about $0.0133 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
hound-mcp CLAUDE.md
Claude Code instructions for tiluckdave/hound-mcp, covering hound mcp — contributor guide, tech stack, project structure, data sources and common commands.
locklens AGENTS.md
Instructions for BARMPlus/locklens, covering agents.md, 项目定位, 当前技术栈, 关键目录 and 审计核心结构.
proof-of-commitment CLAUDE.md
Instructions for piiiico/proof-of-commitment, covering commit, what this is, the thesis, architecture (v0.2) and key decisions.
dependency-fitness-mcp CLAUDE.md
Claude Code instructions for TweedBeetle/dependency-fitness-mcp, covering dependency-fitness-mcp — project memory, the product invariant (the moat — don't erode it), commands, layout and build gotchas (sdk 1.29 + zod v4).
node-modules-inspector AGENTS.md
Instructions for antfu/node-modules-inspector, covering agents.md, layout, commands (from repo root) and conventions.
kkRepo copilot-instructions.md
Copilot instructions for klboke/kkRepo, covering kkrepo copilot instructions, engineering rules, build and test, compatibility workflow and review focus.