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/polos-dev/polos/claude-mdgit clone --depth 1 https://github.com/polos-dev/polosWhat 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 | $0.02140 | $0.02140 |
| Opus 5 | $0.01070 | $0.01070 |
| Sonnet 5 | $0.00428 | $0.00428 |
| Haiku 4.5 | $0.00214 | $0.00214 |
Grade A, and why
polos 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 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.
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 — 254 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md — Polos Developer Guide
Polos is the open-source runtime for AI agents. You write the agent; Polos handles sandboxes, durability, approvals, triggers, and observability.
Docs: https://polos.dev/docs | Repo: https://github.com/polos-dev/polos | Discord: https://discord.gg/ZAxHKMPwFG
Architecture
┌─────────────────────────────────────────────────────────┐
│ polos CLI (Rust) │
│ Single binary embedding orchestrator + UI │
├──────────────────────┬──────────────────────────────────┤
│ Orchestrator (Rust) │ UI (React/TS) │
│ Axum + Tokio │ Vite + Tailwind + Shadcn │
│ PostgreSQL (SQLx) │ localhost:5173 │
│ localhost:8080 │ │
├──────────────────────┴──────────────────────────────────┤
│ Worker (Python or TypeScript SDK) │
│ Executes agents/workflows, connects to orchestrator │
└─────────────────────────────────────────────────────────┘
Orchestrator (orchestrator/) — Core execution engine: durable logs, retries, scheduling, triggers, sandbox management. Rust with Axum, Tokio, SQLx (PostgreSQL).
CLI (server/) — The polos binary. Manages orchestrator and UI processes. Rust with Clap.
UI (ui/) — React 19 dashboard for monitoring agents, workflows, traces. Vite, TypeScript, Tailwind CSS, Shadcn/Radix UI.
Python SDK (sdk/python/) — Agent/workflow definitions, tool system, worker runtime. Pydantic, FastAPI, OpenTelemetry.
TypeScript SDK (sdk/typescript/) — Same capabilities as Python SDK. Fastify, Vercel AI SDK, OpenTelemetry.
Repository Layout
orchestrator/ Rust orchestrator (Axum API server + execution engine)
migrations/ PostgreSQL migration files (0001–0004)
src/ Source code (lib.rs entry point)
tests/ Integration tests (require DB)
server/ Rust CLI binary (embeds orchestrator + UI at build time)
ui/ React/TypeScript dashboard
src/ Components, pages, hooks, API client
sdk/python/ Python SDK (polos-sdk on PyPI)
polos/ Package source
tests/ pytest test suite
sdk/typescript/ TypeScript SDK (@polos/sdk on npm)
src/ Package source
tests/ Test files (*.test.ts)
python-examples/ 20+ example agent projects (Python)
typescript-examples/ 20+ example agent projects (TypeScript)
create-polos-py/ Python project scaffolder (Click CLI)
create-polos-ts/ TypeScript project scaffolder (Clack prompts)
docs/ Documentation site (polos.dev)
DESIGN/ Architecture and design documents
scripts/ Build and dev setup scripts
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 First seen · 254 lines · 2,140 tokens per session scan A 60ef66d475f1
polos CLAUDE.md is an instructions file published in the GitHub repository polos-dev/polos (36 stars, last pushed 6mo ago), licensed Apache-2.0. It adds 2,140 tokens to every session, about $0.0107 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 2026-08-30.
Other instructions, from other repositories
rn-dev-agent AGENTS.md
Instructions for Lykhoyda/rn-dev-agent, covering repository guide for agents, repository map, editing rules, architecture rules and supported node runtimes.
agents CLAUDE.md
Instructions for eloylp/agents, covering claude.md, project overview, directory structure, config model and build & run.
open-bridge AGENTS.md
Instructions for bks-lab/open-bridge, covering the bridge — agent instructions, required reading, session start detection (automatic), theme and agents.
oh-my-multica AGENTS.md
Instructions for xiaohei-info/oh-my-multica, covering agents.md — repository development conventions, hard rules, tdd, errors teach users and definition of done.
designpowers GEMINI.md
Gemini CLI instructions for Owl-Listener/designpowers, covering designpowers, mandatory: welcome sequence first, skills, agents and notes for gemini cli.
kandev CLAUDE.md
Claude Code instructions for kdlbs/kandev, a project described as: AI Kanban & Development Environment. Orchestrate multiple agents, review changes, open PRs. Multi-provider, self-hostable, no telemetry.