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/moonrunnerkc/swarm-orchestrator/agents-mdgit clone --depth 1 https://github.com/moonrunnerkc/swarm-orchestratorWrote 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/moonrunnerkc/swarm-orchestrator/agents-md)<a href="https://agentmods.dev/instructions/moonrunnerkc/swarm-orchestrator/agents-md"><img src="https://agentmods.dev/badge/instructions/moonrunnerkc/swarm-orchestrator/agents-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 | $0.04910 | $0.04910 |
| Opus 5 | $0.02455 | $0.02455 |
| Sonnet 5 | $0.00982 | $0.00982 |
| Haiku 4.5 | $0.00491 | $0.00491 |
Grade E, and why
swarm-orchestrator AGENTS.md scanned grade E with 3 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 2d 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.
Reaches for credential fileshighPrivilege escalation
SSH keys, cloud credentials, git-credentials, .npmrc, /etc/shadow: reading these is how a config file becomes a credential leak.
3. Every tool call passes through the chokepoint. Adding a tool means adding a definition, not a new execution path. The chokepoint decodes before it validates, and only where the tool's own schema says a container was w Downloads and executes remote codehighSupply chain
curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.
3. Every tool call passes through the chokepoint. Adding a tool means adding a definition, not a new execution path. The chokepoint decodes before it validates, and only where the tool's own schema says a container was w Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
3. Every tool call passes through the chokepoint. Adding a tool means adding a definition, not a new execution path. The chokepoint decodes before it validates, and only where the tool's own schema says a container was w Copies of this mod
1 near-identical copy found in the catalogue:
- swarm-orchestrator CLAUDE.md — 100% identical, 0 lines differ
How it starts
The opening of the file, as written. The whole thing — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
Swarm Orchestrator v13: an evidence-first coding agent. Every claim of completed work must resolve to machine-captured evidence in a tamper-evident ledger. Quality gates run and auto-resolve under a ratchet. Local model selection is measured, never guessed. Full rationale lives in docs/build-guide.md; read it before structural work.
Commands
npm run typecheck: tsc, strict, no emitnpm run lint: Biome checknpm run format: Biome format, writenpm test: Vitest, full suitenpm run gates: all of the above in sequence; this is the definition of greennpm run dev: run the CLI from source against a scratch workspace
Run npm run gates before claiming any task complete. Paste the real output. Never summarize gate results from memory.
Architecture Map
src/agent-run.ts: one task start to finish (sandbox, tools, chokepoint, loop, gates). The CLI and every parallel worker call this same function; a worker differs in its directory, its chain, and one optional tool, the peer trail, which the single-agent path is never offered and which leaves its prompt byte-identical.src/core: agent loop. Plan, act, verify. All stochastic inputs (clock, random, model) injected via interfaces, never imported directly.src/tools: read, write, edit, shell, search, list. Every call goes through the chokepoint insrc/tools/chokepoint.ts: ledger record, provenance tag, sandbox enforcement. No tool may bypass it.src/evidence: append-only JSONL ledger, hash chain, content-addressed blob store, evidence DAG, bundle export, embedded verifier, HTML review renderer.src/gates: gate definitions as data, runner, auto-resolve loop with ratchet, escalation.src/providers: the only module allowed to import the Vercel AI SDK. Frontier plus OpenAI-compatible local (Ollama, rapid-mlx). Local endpoint discovery.src/select: hardware probe, static shortlist fit, calibration micro-eval, bandit reward log.src/workers: phase 6 scale-out. Git worktree per worker, each running the ordinary loop fromsrc/agent-run.ts, and a merge queue that lands them sequentially under the ratchet. Nothing here is imported by the single-agent path.src/tui: Ink single-screen UI. Renders exclusively from ledger projections.src/config: Zod-validated swarm.toml, zero-config defaults.
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.
- 2d ago Changed · +2 lines · +320 tokens per session 2da03f7b0d80
- 6d ago First seen · 68 lines · 4,590 tokens per session scan E 9742d1c0cc28
swarm-orchestrator AGENTS.md is an instructions file published in the GitHub repository moonrunnerkc/swarm-orchestrator (106 stars, last pushed yesterday), licensed ISC. It adds 4,910 tokens to every session, about $0.0246 per session on Opus 5. A static security scan graded it E with 3 findings (reaches for credential files, downloads and executes remote code, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
anatomia AGENTS.md
Instructions for anatomia-dev/anatomia, covering anatomia-workspace, commands and conventions.
vscode design-philosophy.instructions.md
VS Code design philosophy — the shared Values→Principles→Moves vocabulary for reasoning about UI in design terms rather than raw pixels. Use when creating, editing, or reviewing any visual surface (CSS, DOM, theming, icons, motion). Name the value/principle before reaching for a token.
deepagents AGENTS.md
AGENTS.md instructions for langchain-ai/deepagents, covering global development guidelines for the deep agents monorepo, corridor security analysis, development workflow, suppressing ruff rules and pr conventions.
retry-axios GEMINI.md
Gemini CLI instructions for JustinBeckwith/retry-axios, covering project: retry-axios, project overview, building and running, dependencies and commands.
huly-mcp CLAUDE.md
Claude Code instructions for dearlordylord/huly-mcp, covering project instructions, design principle: llm-first api, project harness (copy to new projects), package manager and verification.
markdy-com copilot-instructions.md
Copilot instructions for HoangYell/markdy-com, a project described as: 🎬 Open-source animated architecture diagrams as code. Diagram-native DSL, 17 layout engines, Web Animations API, and MCP server for AI agents.