What is Operator-Use CLAUDE.md?

Repository instructions for Operator, a Python framework that runs stateful AI agents across multiple providers, tools, sessions, extensions, and messaging channels. They describe its four main layers and where the code for each layer lives.

Navigating the repository, understanding the architecture, and working on inference, sessions, extensions, automation, scheduling, or connections such as Telegram, Slack, WebSocket, and email.

They give an agent a map of the system and its boundaries, making it easier to find the right code and preserve the separation between engine, agent, runtime, and gateway responsibilities.

Operator-Use CLAUDE.md — Instructions file for coding agents by Jeomon: Repository instructions for Operator, a Python framework that runs…

Install

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.

agentmods

npx agentmods add instructions/jeomon/operator-use/claude-md

Clone the repo

git clone --depth 1 https://github.com/Jeomon/Operator-Use

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/jeomon/operator-use/claude-md.svg)](https://agentmods.dev/instructions/jeomon/operator-use/claude-md)

Your own site

<a href="https://agentmods.dev/instructions/jeomon/operator-use/claude-md"><img src="https://agentmods.dev/badge/instructions/jeomon/operator-use/claude-md.svg" alt="Measured on agentmods" height="20"></a>

Per session 1,675 This file is loaded in full into every session.

When invoked 1,675 The same file — it is already loaded in full.

Security scan A 0 findings. Scan, not verified.

Origin original No closer match found in the catalogue.

Token cost

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	$0.01675	$0.01675
Opus 5	$0.00838	$0.00838
Sonnet 5	$0.00335	$0.00335
Haiku 4.5	$0.00168	$0.00168

Measured 3d ago against content hash 7645efd99bae, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

Operator-Use 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 3d 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.

CLAUDE.md · 142 lines

How it starts

The opening of the file, as written. The whole thing — 142 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Operator — CLAUDE.md

Top-level guidance for Claude Code when working on this repository.

Project Overview

Operator is a stateful Python AI agent harness. It wraps a multi-provider LLM inference layer with session persistence, context compaction, an extension/package system, sandbox enforcement, browser/computer automation, and a gateway that connects the agent to multiple messaging channels simultaneously (Telegram, Discord, Slack, WebSocket, Email, Twitch, stdio).

Four layers build on each other:

Engine (operator_use/engine/) — raw LLM streaming loop, tool execution, abort signal. No knowledge of sessions or extensions.
Agent (operator_use/agent/) — turn orchestration, retry, compaction scheduling, extension event fan-out. Implements ExtensionContext.
Runtime (operator_use/runtime/) — session lifecycle, slash-command dispatch, gateway/cron/subagent wiring.
Gateway (operator_use/gateway/) — channel adapters, async message bus, per-session agent routing.

Repository Layout

operator_use/
  agent/         engine/        session/       runtime/
  extension/     hooks/         resource/      gateway/
  inference/     settings/      package/       builtins/
  compaction/    subagent/      cron/          process/
  mcp/           acp/           auth/          tool/
  skill/         commands/      message/       prompt/
  browser/       computer/      peer/          team/
  workflow/      knowledge/     sandbox/       diagnostics/
  memory/        bus/           console/       templates/
tests/
docs/            ← read these before exploring code

Find detailed documentation before reading source:

Need to understand…	Read
Full system — channels → gateway → agent → engine → session	`docs/architecture.md`
Turn flow, retry, compaction scheduling	`docs/agent.md`
Guardrail interface, loading, loop detection	`docs/guardrails.md`
LLM loop, tool execution modes, queues	`docs/engine.md`
Session JSONL, branching, reconstruction	`docs/session.md`
Extension loading, `api.config`, dispatch	`docs/extensions.md`
Package install, `operator.json`, wiring	`docs/packages.md`
Event types, hook registration	`docs/hooks.md`
Channels, message bus, stream phases	`docs/gateway.md`
Models, providers, auth	`docs/inference.md`
Tool interface, execution modes	`docs/tool.md`
SKILL.md format, discovery	`docs/skill.md`
Slash commands, builtin commands	`docs/commands.md`
Credentials, OAuth, token refresh	`docs/auth.md`
Named profiles, AGENT.md format	`docs/profiles.md`
Browser automation via CDP	`docs/browser.md`
Desktop computer control	`docs/computer.md`
Sandbox policy, OS-level isolation	`docs/sandbox.md`
Knowledge base, document injection	`docs/knowledge.md`
Multi-agent teams, mailboxes	`docs/team.md`
Python workflows, DSL globals	`docs/workflow.md`
ACP transports, server/client	`docs/acp.md`

Read the full file on GitHub · 142 lines

Changes

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.

3d ago First seen · 142 lines · 1,675 tokens per session scan A 7645efd99bae

Subscribe to this mod's changes

Operator-Use CLAUDE.md is an instructions file published in the GitHub repository Jeomon/Operator-Use (40 stars, last pushed 2mo ago), licensed MIT. It adds 1,675 tokens to every session, about $0.0084 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.

ai-agents browser-use computer-use openclaw personal-assistant

Operator-Use CLAUDE.md

Getting it into your agent

Wrote this? Show the measurements

What it costs to keep this loaded

Grade A, and why

Nothing flagged

How it starts

Operator — CLAUDE.md

Project Overview

Repository Layout

Quick Navigation

What this file has done since we first saw it

Other instructions, from other repositories

SmolVM AGENTS.md

SmolVM CLAUDE.md

lobu AGENTS.md

lobu CLAUDE.md

librefang AGENTS.md

openclaw CLAUDE.md