Borrowing it
Nothing to install: this file belongs to arc-mcp/arc-1. 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/arc-mcp/arc-1/main/AGENTS.mdgit clone --depth 1 https://github.com/arc-mcp/arc-1Wrote 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/arc-mcp/arc-1/agents-md)<a href="https://agentmods.dev/instructions/arc-mcp/arc-1/agents-md"><img src="https://agentmods.dev/badge/instructions/arc-mcp/arc-1/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/arc-mcp/arc-1/agents-md"><img src="https://agentmods.dev/badge/instructions/arc-mcp/arc-1/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 | $0.15504 | $0.15504 |
| Opus 5.5 | $0.06202 | $0.06202 |
| Sonnet 5.5 | $0.03101 | $0.03101 |
| Haiku 4.5 | $0.01550 | $0.01550 |
Grade A, and why
arc-1 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 today.
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 — 425 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
Guidance for AI coding agents (Claude Code, Codex, Cursor, Copilot, …) working in this repository.
Single source of truth — CLAUDE.md imports this file. Keep it terse: task→files + ≤1 gotcha per
row; verbose details and live-verified behaviors live in docs/dev-guide.md
(not auto-loaded — read the matching row there when working on one of these tasks).
Project Overview
ARC-1 is a TypeScript MCP (Model Context Protocol) server for SAP ABAP Development Tools (ADT).
It provides 12 intent-based tools (SAPRead, SAPSearch, SAPWrite, SAPActivate, SAPNavigate, SAPQuery,
SAPTransport, SAPGit, SAPContext, SAPLint, SAPDiagnose, SAPManage) for Claude and other MCP clients.
Distributed as npm package (arc-1) and Docker image (ghcr.io/arc-mcp/arc-1).
Design Principles
- Centralized admin control — managed service; server-wide safety ceiling (
allowWrites, package allowlists, SQL/data/transport/Git gates, deny actions); every call audited; per-user scopes restrict, never expand. - Per-user SAP identity by default — principal propagation maps each MCP user to their own SAP user. ADR-0007 permits only an explicit mutation-free multi-target Basic exception, labeled shared and never used as PP fallback.
- Token-efficient tools — 12 intent tools vs 200+ endpoints, with schema payload guarded by CI budgets; hyperfocused mode = 1 tool (~200 tokens); method-level surgery + context compression keep mid-tier LLMs viable.
- BTP-native deployment — Destination Service, Cloud Connector, XSUAA OAuth, BTP Audit Log; also Docker/npm/stdio.
- Multi-client, vendor-neutral — XSUAA OAuth + Entra ID OIDC + API key coexist; one instance serves Claude, Copilot Studio, VS Code, Gemini CLI, Cursor.
- Safe defaults, opt-in power — read-only by default; free SQL blocked; package allowlist defaults to
$TMP; everything forbidden until the admin allows it. - Single target by default; explicit BTP exceptions — ADR-0006 permits default-off discovered pinned/aggregate targets; ADR-0007 keeps shared Basic mutation-free and one-instance. ADR-0010 permits only scoped CLAS/INTF writes and whole-owner activation on pinned strict-PP routes with destination restrictions and exact verified write grants. Aggregate/Copilot remain mutation-free. Broader writes or another discovery/auth model require a new ADR/security review.
- ABAP execution and RFC belong in extensions — no built-in report/class execution, direct RFC, or installed/temporary ABAP helpers. Core uses existing SAP APIs and TypeScript; user ABAP development and dedicated ATC/Unit workflows retain their gates. See ADR-0009.
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.
- today Changed · +46 tokens per session cbdb18be11f0
- 3d ago Changed · +1 lines · +34 tokens per session 249f2af7d15e
- 4d ago Changed · +293 tokens per session 82f1f5766ce8
- 5d ago Changed · +1 lines · +82 tokens per session d81cfea17980
- 7d ago Changed · +1 lines · +85 tokens per session 1a1e77ab2a85
- 8d ago Changed · +1 lines · +59 tokens per session 79ebafa8bb83
- 9d ago Changed · +4 lines · +134 tokens per session ed0f22097149
- 12d ago Changed · +5 lines · +186 tokens per session 822d91267cc3
- 14d ago Changed · +4 lines · +333 tokens per session f9f828e6b640
- 15d ago First seen · 408 lines · 14,252 tokens per session scan A f1a2952833b8
arc-1 AGENTS.md is an instructions file published in the GitHub repository arc-mcp/arc-1 (239 stars, last pushed today), licensed MIT. It adds 15,504 tokens to every session, about $0.0620 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).
deno copilot-instructions.md
Copilot instructions for denoland/deno, covering deno development guide for github copilot, network access, high level overview, key directories and key files to understand first.