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/aleksbuss/orchestra/claude-mdgit clone --depth 1 https://github.com/aleksbuss/orchestraWrote 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/aleksbuss/orchestra/claude-md)<a href="https://agentmods.dev/instructions/aleksbuss/orchestra/claude-md"><img src="https://agentmods.dev/badge/instructions/aleksbuss/orchestra/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 | $0.07967 | $0.07967 |
| Opus 5 | $0.03984 | $0.03984 |
| Sonnet 5 | $0.01593 | $0.01593 |
| Haiku 4.5 | $0.00797 | $0.00797 |
Grade A, and why
orchestra CLAUDE.md scanned grade A with 2 findings against 26 rules in 11 categories โ prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency โ 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
const res = await fetch(safeUrl, { signal: AbortSignal.timeout(5000) }); Runs shell commandslowCapability
Expected in a hook, worth knowing in a rule or an instructions file.
| Wrapping an external CLI for the agent, or reaching for `child_process.exec` in a tool | [`docs/references/graphify-integration-adr.md`](docs/references/graphify-integration-adr.md) | Tool-vs-skill-vs-MCP decision + th How it starts
The opening of the file, as written. The whole thing โ 270 lines โ stays where its author put it; the contents beside it link to each section on GitHub.
Orchestra Swarm Engine
๐ค System Prompt for AI Assistants
You are an expert AI Full-Stack Software Engineer with deep expertise in Next.js 15 (App Router), TypeScript, Zustand, and Vercel AI SDK. When working on this repository, you must strictly follow these rules:
- Write robust, self-healing, and defensive code.
- Avoid introducing technical debt. If a pattern exists in the codebase (e.g.,
safeWriteFile), you MUST use it rather than reinventing standard Node.js libraries. - Prefer explicit TypeScript typing over
anyor implicit inference. - Do not remove existing comments unless explicitly refactoring the commented logic.
- Consult
POST_MORTEMS.mdbefore refactoring core logic (especially SSE streams, MoA, or file storage) to avoid repeating known historical bugs.
๐ How this document is organized (read before adding to it)
This file is Level 1: it loads into every single session, so every line spends context budget that the task itself needs. Detailed contracts live in Level 2 (docs/references/*.md) and are loaded on demand via the trigger tables below.
Level 1 (this file) holds only
| Type | Example |
|---|---|
| Core commands | npm run verify:strict |
| Non-negotiable rules | every user-supplied path goes through assertPathInside |
| Copy-paste code patterns | the SSRF-guard call shape |
| Directory map | feature โ file |
| Trigger indexes | pointers into Level 2 |
Level 2 (docs/references/) holds
Full contract text, per-PM rationale, historical decisions, shipped-track narrative, edge cases, seam plans.
When you are asked to record something here
- Is it a rule that applies to most sessions? โ Level 1, as one imperative line.
- Is it rationale, history, an edge case, or a shipped-track narrative? โ Level 2 file, plus a trigger-table row here if a new trigger appears.
- Is it already enforced by a CI gate? โ do not restate the rule in prose; the test is the rule. Add it to the CI-gate table below.
- Is the canonical source another document (
POST_MORTEMS.md,docs/*.md, git history)? โ link it; do not copy it.
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 ยท +1 lines ยท +134 tokens per session 3661e23a208f
- 5d ago First seen ยท 269 lines ยท 7,833 tokens per session scan A cb41ed3996ac
orchestra CLAUDE.md is an instructions file published in the GitHub repository aleksbuss/orchestra (2 stars, last pushed yesterday), licensed MIT. It adds 7,967 tokens to every session, about $0.0398 per session on Opus 5. A static security scan graded it A with 2 findings (makes network calls, runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
playbooks CLAUDE.md
Instructions for playbooks-ai/playbooks, covering claude.md, project overview, core architecture, framework components and key architectural insights.
agent-claim-network AGENTS.md
AGENTS.md instructions for FTShare-Lab/agent-claim-network, covering agents.md, ้กน็ฎๆฆ่ง, ๅฝๅๆถๆ, rust ่ง่ and ๆฐๆฎไธๅฎๅ จ่พน็.
multi_mcp CLAUDE.md
Claude Code instructions for religa/multi_mcp, covering claude.md, project overview, current status, development commands and type checking (required before commits).
treelang AGENTS.md
AGENTS.md instructions for cs0lar/treelang, covering repository guidelines, project structure & module organization, build, test, and development commands, coding style & naming conventions and testing guidelines.
CommonGround AGENTS.md
Instructions for Intelligent-Internet/CommonGround, covering repository guidelines, repository role, documentation priority, code structure and common commands.
Orca CLAUDE.md
Claude Code instructions for junkyard22/Orca, covering orca agent instructions, contract check, miranda architecture lock and where miranda is wired today.