Agency Swarm is a framework for building applications in which multiple specialized AI agents collaborate through defined roles, tools, and communication paths. Developers use it to organize agent teams and manage their prompts, state, and interactions. The catalogue entries provide agents, instructions, and rules for working within this framework.
Borrowing it
Nothing to install: this file belongs to VRSEN/agency-swarm. 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/VRSEN/agency-swarm/main/AGENTS.mdgit clone --depth 1 https://github.com/VRSEN/agency-swarmWrote 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/vrsen/agency-swarm/agents-md)<a href="https://agentmods.dev/instructions/vrsen/agency-swarm/agents-md"><img src="https://agentmods.dev/badge/instructions/vrsen/agency-swarm/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/vrsen/agency-swarm/agents-md"><img src="https://agentmods.dev/badge/instructions/vrsen/agency-swarm/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 | $0.01036 | $0.01036 |
| Opus 5.5 | $0.00414 | $0.00414 |
| Sonnet 5.5 | $0.00207 | $0.00207 |
| Haiku 4.5 | $0.00104 | $0.00104 |
Grade A, and why
agency-swarm 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 15d 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 — 47 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agency Swarm Repository Rules
Core principle, in the maintainer's words: "Agency Swarm should remain a focused orchestration layer over the OpenAI Agents SDK, not grow into a duplicate agent runtime. Whenever possible, use the OpenAI Agents SDK instead of reimplementing its behavior." Check every change against it.
Multi-step procedures live in .claude/skills/: review, release, sdk-upgrade. Load the matching skill before you start one of those.
1. The OpenAI Agents SDK comes first
1.1 Before you write or change code, find how the openai-agents version pinned in pyproject.toml already does it: read its installed source in .venv and references/openai-agents-python at that tag (clone it there if missing). If the SDK has it, call it, configure it or subclass it. Never copy its logic or write a parallel version.
1.2 Write custom code only when the SDK has no equivalent. The pull request names the SDK feature you checked and why it does not fit.
1.3 Code that wraps or replaces an SDK abstraction (session, history items, streaming events, model settings, tracing, guardrails, handoffs) names the SDK class it mirrors in its docstring and supports every feature of that class, or lists each unsupported feature with the reason. A custom session layer once missed SDK compaction for months; this rule exists so that does not repeat.
1.4 When the SDK adds a feature our code duplicates, delete our code and use the SDK's in the same change. Keep our version only for a concrete reason written in the pull request.
1.5 Third-party and vendor integrations stay outside this repository. Decline vendor-pitch issues and pull requests, point to the extension seam, and ship at most a docs recipe.
2. Moving fast and public actions
2.1 Backward compatibility is not a goal by itself. Make the better change, bump the version, and document what changed in the release notes and migration guide. Do not add compatibility shims. Escalate to the maintainer only a change that breaks users who cannot upgrade, such as an old deployed template.
2.2 Get the maintainer's approval of the exact text or action before any public action on GitHub: comments, reviews, issues, pull requests, merges, closes, labels, pushes to shared branches, tags, releases and package uploads. Prepare drafts. Exception: a short reply that corrects a clearly wrong automated comment.
2.3 No AI attribution in commits or pull requests: no AI Co-Authored-By trailers, no "Generated with" footers.
2.4 Code for a deployed agency ships through the local clone and a push to GitHub; the hosting platform rebuilds from the push. Never edit code in the platform editor.
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.
- 15d ago Changed · -30 lines · -664 tokens per session 8600c14ccd59
- 18d ago Changed · +2 lines · +95 tokens per session fffc8c0a9d71
- 19d ago Changed · +4 lines · +108 tokens per session bd5fa928c081
- 1mo ago Changed · -4 lines · -81 tokens per session 380e6450aa8a
- 1mo ago Changed · +5 lines · +192 tokens per session 914cdba3e8be
- 1mo ago Changed · +5 lines · +153 tokens per session 13d5edc7a6a7
- 1mo ago Changed · +1 lines · +49 tokens per session 1b7691119223
- 1mo ago Changed · +1 lines · +45 tokens per session 571fdea2e8e6
- 1mo ago First seen · 63 lines · 1,139 tokens per session scan A 45ceb758de8d
agency-swarm AGENTS.md is an instructions file published in the GitHub repository VRSEN/agency-swarm (4,617 stars, last pushed today), licensed MIT. It adds 1,036 tokens to every session, about $0.0041 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.