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/untra/operator/claude-mdgit clone --depth 1 https://github.com/untra/operatorWhat 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 | $0.03628 | $0.03628 |
| Opus 5 | $0.01814 | $0.01814 |
| Sonnet 5 | $0.00726 | $0.00726 |
| Haiku 4.5 | $0.00363 | $0.00363 |
Grade A, and why
operator 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 — 309 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md - operator
Project Overview
operator is a Rust TUI application for orchestrating Claude Code agents across multi-project codebases. It manages ticket queues, launches agents, tracks progress, and provides notifications.
Tech Stack
- Language: Rust (core), TypeScript (
ui/,webcomponents/,vscode-extension/) - TUI: ratatui (crossterm backend); tokio async runtime
- Server surfaces: axum REST API + utoipa OpenAPI (
src/rest/), MCP server (src/mcp/), ACP (src/acp/) - Web: Vite/React SPA in
ui/, sharedwebcomponents/package, thin VS Code extension webview - Notifications: mac-notification-sys (macOS), notify-rust (Linux), webhooks
- Config: config crate (TOML); File Watching: notify crate
Code Style
Aim for functional software development with a focus on stateless, single responsibility focus.
Minimize use of comments; they should be terse and used judiciously, ideally one sentence tops.
Data types come from rust; typescript and docs binds are generated from low-level rust types annotated with comments that embed as descriptions into configuration and reference files.
Favor falsey defaults ; lets aim not to enforce default=true or some other javascript-truthy default value.
Plans & Specs Location
Write superpowers plans to superpowers/plans/ and design specs to superpowers/specs/ (repo root, not hosted).
Development Standards
- Ask, don't assume. If something is unclear, ask before writing into a corner. Never make silent assumptions about intent, architecture or requirements.
- Consider the simplest solution first. First attempt the simplest approach that could work, then consider abstractions and flexibility with regards to similar implementations.
- Avoid changing unrelated code. If a file or function is not part of the current task, do not modify it.
- Flag uncertainty explicitly. If you are not confident about an approach or technical detail, say so before proceeding.
- I am open to ideas on better approaches or strategies. Speak up and suggest a better course if I am off-course from a better solution.
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 · 309 lines · 3,628 tokens per session scan A 69cd82081266
operator CLAUDE.md is an instructions file published in the GitHub repository untra/operator (37 stars, last pushed 5d ago), licensed MIT. It adds 3,628 tokens to every session, about $0.0181 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
multi-clone-code CLAUDE.md
Instructions for maflot/multi-clone-code, covering multi-claude orchestra, architecture, key design decisions and working on this repo.
tmux-kanban CLAUDE.md
Instructions for linwk20/tmux-kanban, a project described as: An AI-native, lightweight web kanban board for tmux sessions and terminal agents.
obsidian-agent-client AGENTS.md
AGENTS.md instructions for RAIT-09/obsidian-agent-client, covering agent client plugin - llm developer guide, architecture, data flow, acp event flow (single path) and permission flow.
veritas-kanban AGENTS.md
Instructions for BradGroux/veritas-kanban, covering agents.md — canonical agent instructions for veritas kanban, runtime requirements, repository layout, essential commands and install.
memex GEMINI.md
Instructions for STiFLeR7/memex, covering memex, agent instructions, execution mode, non-negotiables and what is in scope.
jjj AGENTS.md
Instructions for doug/jjj, covering agent instructions, before you write anything, read this first, house rules and verifying changes to this repository.