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/2389-research/building-multiagent-systems/claude-mdgit clone --depth 1 https://github.com/2389-research/building-multiagent-systemsWrote 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/2389-research/building-multiagent-systems/claude-md)<a href="https://agentmods.dev/instructions/2389-research/building-multiagent-systems/claude-md"><img src="https://agentmods.dev/badge/instructions/2389-research/building-multiagent-systems/claude-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 5.1 | $0.00878 | $0.00878 |
| Opus 5 | $0.00439 | $0.00439 |
| Sonnet 5 | $0.00176 | $0.00176 |
| Haiku 4.5 | $0.00088 | $0.00088 |
Grade A, and why
building-multiagent-systems CLAUDE.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 6d 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 — 112 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Building Multi-Agent Systems Plugin
Overview
This plugin provides comprehensive architecture patterns for multi-agent systems where AI agents coordinate to accomplish complex tasks using tools. Language-agnostic and applicable across TypeScript, Python, Go, Rust, and other environments.
Skill Included
building-multiagent-systems
Trigger keywords: multi-agent, orchestrator, coordination, sub-agents, delegation, parallel work, sequential pipeline, fan-out, map-reduce
When to use:
- Designing systems where multiple AI agents coordinate
- Implementing orchestrators that spawn sub-agents
- Building systems with parallel or sequential agent workflows
- Coordinating shared resources across agents
- Managing agent lifecycle and state
What it does:
- Asks six discovery questions about requirements and constraints
- Presents foundational patterns (event-sourcing, hierarchical IDs, state machines)
- Recommends coordination pattern (fan-out/fan-in, pipeline, delegation, queue, map-reduce, peer collaboration)
- Guides tool coordination design (permissions, locking, rate limiting, caching)
- Enforces cascading cleanup to prevent orphaned agents
- Provides production-ready implementation examples
Patterns
Six Coordination Patterns
- Fan-Out/Fan-In - Parallel independent work with batching
- Sequential Pipeline - Multi-stage transformations with checkpointing
- Recursive Delegation - Hierarchical task breakdown with depth limits
- Work-Stealing Queue - Large batches with load balancing
- Map-Reduce - Cost optimization with cheap map, smart reduce
- Peer Collaboration - LLM council for bias reduction
Foundational Patterns
- Event-Sourcing - All state changes as events for audit trails and replay
- Hierarchical IDs - Encode delegation hierarchy (e.g.,
session.1.2) - Agent State Machines - Explicit states with invalid transition errors
- Communication Mechanisms - EventEmitter, Channels, Async/Await
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.
- 6d ago First seen · 112 lines · 878 tokens per session scan A 9a588179b683
building-multiagent-systems CLAUDE.md is an instructions file published in the GitHub repository 2389-research/building-multiagent-systems (5 stars, last pushed 2mo ago), licensed MIT. It adds 878 tokens to every session, about $0.0044 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 2026-08-31.
Other instructions, from other repositories
Orra-mcp CLAUDE.md
Instructions for bjornj12/Orra-mcp, covering orra mcp development, project, architecture, testing and structure.
research-os AGENTS.md
Instructions for lxinfei5/research-os, covering researchos — constitution, §0 what this is, §1 directory discipline, §2 half-life (innovation 1) — memory design and §3 behavior pillars (innovations 2–5).
sema AGENTS.md
AGENTS.md instructions for emergent-wisdom/sema, covering repository agent guide, development workflow, vocabulary changes, vocabulary review and the improvement loop, per pattern.
sema CLAUDE.md
Claude Code instructions for emergent-wisdom/sema: Read and follow AGENTS.md. It is the canonical repository-level workflow for development, vocabulary changes, deployment boundaries, and verification.
dibs CLAUDE.md
Claude Code instructions for Agenxy/dibs, a project described as: Keeps your agents in the loop about each other. See what the fleet is doing, then message, sync and send files between agents. Dibs reports; it never acts.
lobsterroll CLAUDE.md
Instructions for onEnterFrame/lobsterroll, covering claude.md, what this is, commands, install dependencies and build all packages (must build shared → db → api in order).