Terminator is a computer-use MCP server that lets AI assistants control desktop applications and browsers through their existing sessions. Developers use it to automate tasks such as managing cloud resources, checking service logs, and testing applications. Its catalogue instruction, rule, and command support agent workflows that operate the desktop.
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/mediar-ai/terminator/claude-mdgit clone --depth 1 https://github.com/mediar-ai/terminatorWrote 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/mediar-ai/terminator/claude-md)<a href="https://agentmods.dev/instructions/mediar-ai/terminator/claude-md"><img src="https://agentmods.dev/badge/instructions/mediar-ai/terminator/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 | $0.00504 | $0.00504 |
| Opus 5 | $0.00252 | $0.00252 |
| Sonnet 5 | $0.00101 | $0.00101 |
| Haiku 4.5 | $0.00050 | $0.00050 |
Grade A, and why
terminator 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 5d 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.
What it actually says
Terminator - AI-Native GUI Automation
Open-source desktop automation (MIT). Gives AI hands to control any app. Windows-only. Mediar AI | $2.8M seed | mediar.ai
Installation
npx @mediar-ai/cli --help # Run without install
npm i -g @mediar-ai/cli # Or install globally
Release
CRITICAL: Use terminator CLI only (syncs versions across workspace).
npx @mediar-ai/cli release # Bump patch → tag → push (triggers CI/CD)
# Or with globally installed CLI:
terminator release
Never manually edit versions in package.json or Cargo.toml.
Git tags trigger: publish-npm.yml (npm), publish-mcp.yml (npm), ci-wheels.yml (PyPI).
Development
git clone https://github.com/mediar-ai/terminator && cd terminator
cargo build && cargo test
cargo fmt && cargo clippy
Critical Patterns
TypeScript workflows:
- Use
context.data = {...}in final step for MCP integration - See
examples/simple_notepad_workflow/for best practices - If CLI shows "(no parser)" → check
workflow_typescript.rs:379-398
Security:
- ❌ NEVER use
#idselectors (non-deterministic across machines) - ❌ Don't commit credentials
- ✅ Use
role:Type && name:Nameornativeidselectors
MCP Debugging
Logs: %LOCALAPPDATA%\claude-cli-nodejs\Cache\*\mcp-logs-terminator-mcp-agent\*.txt
Enable: Set LOG_LEVEL=debug and RUST_BACKTRACE=1 in MCP config.
Packages
- npm:
@mediar-ai/terminator(2 Windows platform packages),terminator-mcp-agent,@mediar-ai/workflow - PyPI:
terminator - crates.io:
terminator-rs
Repo: crates/terminator*, packages/terminator-*/workflow
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.
- 5d ago First seen · 59 lines · 504 tokens per session scan A b65c45748c84
terminator CLAUDE.md is an instructions file published in the GitHub repository mediar-ai/terminator (1,629 stars, last pushed 3mo ago), licensed MIT. It adds 504 tokens to every session, about $0.0025 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
openscience CLAUDE.md
Claude Code instructions for synthetic-sciences/openscience, covering claude.md: openscience, project overview, repository structure, development and run cli in dev mode.
omnigent copilot-instructions.md
Copilot instructions for omnigent-ai/omnigent, covering copilot code review instructions, e2e test requirement, backend test coverage and frontend test coverage.
openscience AGENTS.md
AGENTS.md instructions for synthetic-sciences/openscience, covering style guide, avoid let statements, avoid else statements, prefer single word naming and testing.
omnigent AGENTS.md
AGENTS.md instructions for omnigent-ai/omnigent, covering agent guidance, committing, local development shortcuts, pull requests and finishing a task.
Aether AGENTS.md
AGENTS.md instructions for Zhou-Shilin/Aether, a project described as: A stunning, localized, general-purpose AI Agent for Android, iOS & Mac. | Aether 扶摇.
omnigent CLAUDE.md
Claude Code instructions for omnigent-ai/omnigent, a project described as: Omnigent is an open-source AI agent framework and meta-harness: orchestrate Claude Code, Codex, Cursor, Pi, and custom agents — swap harnesses without rewriting, enforce policies and sandboxing, and collaborate in real time from any device.