Borrowing it
Nothing to install: this file belongs to howells/arc. 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/howells/arc/main/AGENTS.mdgit clone --depth 1 https://github.com/howells/arcWrote 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/howells/arc/agents-md)<a href="https://agentmods.dev/instructions/howells/arc/agents-md"><img src="https://agentmods.dev/badge/instructions/howells/arc/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/howells/arc/agents-md"><img src="https://agentmods.dev/badge/instructions/howells/arc/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 | $0.01113 | $0.01113 |
| Opus 5.5 | $0.00445 | $0.00445 |
| Sonnet 5.5 | $0.00223 | $0.00223 |
| Haiku 4.5 | $0.00111 | $0.00111 |
Grade A, and why
arc 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 1mo 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 — 49 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Arc
The full arc from idea to shipped code, shipped as a plugin for Claude Code, Codex, and Cursor. CONTEXT.md is the canonical product definition, domain language, and operating boundary - read it before changing workflow behaviour. Treat this file as contributor guidance, not a competing definition.
Shape
- Each skill in
skills/<name>/SKILL.mdis one/arc:<name>command, routed by a thincommands/<name>.md. Never fork behaviour between the two. - Commands:
/arc:ideate,/arc:review,/arc:implement,/arc:testing,/arc:audit,/arc:improve,/arc:refactor,/arc:commit,/arc:release,/arc:launch,/arc:vision. - Canonical flow is
ideate -> (review) -> implement.audit,improve,commit,release, andrefactorare cross-cutting and available anytime. - Skills draw on
agents/(subagents),references/(domain knowledge, catalogued inreferences/index.md),rules/,disciplines/(methodologies), andtemplates/(output structures). skills/using-arc/is the session control plane;skills/detail/is internal toimplement. Neither has a command router.- Site work is isolated under
site/with its own Next.js scripts. rules/is Arc's internal corpus, loaded selectively by workflows. It's shipped in the plugin payload but never injected as always-on Cursor rules -tests/test-cursor-plugin.shenforces that.
Gotchas
- Rebuild the Codex mirror after any change to
skills/,agents/,references/,rules/,disciplines/,templates/,scripts/, orcommands/. Runpnpm build:codex. The pre-commit hook andtests/test-codex-mirror.shboth fail on a stale mirror, and the failure names the diffing file rather than the real cause. plugins/arc/is generated. Never edit it by hand -scripts/build-codex-plugin.shoverwrites it wholesale.- Version bumps touch five manifests. Use
bash scripts/bump-version.sh <version>, never a manual edit: it updatespackage.json,.claude-plugin/,.codex-plugin/,.cursor-plugin/(plugin and marketplace), and the generatedplugins/arc/.codex-plugin/. Verify withbash scripts/bump-version.sh --check. A post-commit hook auto-bumps the patch version and amends content commits. - Adding a skill needs more than a
SKILL.md. It also needs acommands/<name>.mdrouter, a.agents/skills/<name>symlink, and entries in the plugin manifests -tests/test-skill-loading.shhardcodes the expected skill list and fails on any unexpected directory. - Agents are not skills. Every file in
agents/*/*.mdkeeps its own<arc_runtime>and unconditionally loadsreferences/subagent-safety.md. That's deliberate: agents are dispatched as subagents and skipusing-arcvia its<SUBAGENT-STOP>block, so those local blocks are their only path grounding and injection defence.tests/test-xml-tags.shenforces tag containment for agents but not for skills, for the same reason. - Skills must stay self-sufficient.
audit,improve, andrefactorcarrycontext: forkand run isolated; command routers also invoke skills cold. Don't centralise a skill's<arc_runtime>intousing-arc- path resolution would silently break. - Adding a required reference to a skill means the file must exist and the skill must say when to read it.
tests/test-agents-and-refs.shchecks the first half.
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.
- 1mo ago First seen · 49 lines · 1,113 tokens per session scan A 672cf26fd5d6
arc AGENTS.md is an instructions file published in the GitHub repository howells/arc (25 stars, last pushed 1mo ago), licensed MIT. It adds 1,113 tokens to every session, about $0.0045 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 2026-08-30.
Other instructions, from other repositories
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode copilot-instructions.md
Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).
vscode model-prompts.instructions.md
Model-specific prompt authoring and registry guidelines.