terminal-use

terminal-use is a skill for Claude Code, Codex from HLH2023/terminal-use-mcp. It costs 26 tokens per session (9,550 once invoked), scanned C, original, MIT.

A terminal-control guide for interacting with programs that require keyboard input, such as text-based user interfaces.

In plain words
What is it for?
Taking snapshots of terminal sessions, sending keyboard actions, waiting for changes, choosing a terminal provider, and applying safety checks.
Why use it?
It helps an agent operate interactive terminal programs through a repeated view, action, and wait process instead of treating them like ordinary shell commands.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one. Also seen: mentions Claude Code; mentions Codex; mentions Gemini CLI.

Not installable: its command points at a path on the author’s own machine, so it runs nowhere else. The line is /home/user/project.

Good fit Taking snapshots of terminal sessions, sending keyboard actions, waiting for changes, choosing a terminal provider, and applying safety checks.

Compare 6 skills from other repositories ↓
Install

Getting it into your agent

There is no command for this one: it runs only inside a plugin, and the catalogue could not identify which plugin ships it. The source is linked below.

Made for: Claude Code, Codex.

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.

agentmods badge for terminal-use

README.md
[![agentmods](https://agentmods.dev/badge/skills/hlh2023/terminal-use-mcp/terminal-use/github.svg)](https://agentmods.dev/skills/hlh2023/terminal-use-mcp/terminal-use)
Your own site
<a href="https://agentmods.dev/skills/hlh2023/terminal-use-mcp/terminal-use"><img src="https://agentmods.dev/badge/skills/hlh2023/terminal-use-mcp/terminal-use/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for terminal-use

Your own site · 80×15
<a href="https://agentmods.dev/skills/hlh2023/terminal-use-mcp/terminal-use"><img src="https://agentmods.dev/badge/skills/hlh2023/terminal-use-mcp/terminal-use.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 26 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 9,550 The whole file, excluding the scripts and references it only reads on demand.
Security scan C 1 finding. A grade says what 26 rules found in the file — not that it is safe.
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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00026 $0.09550
Opus 5 $0.00013 $0.04775
Sonnet 5 $0.00005 $0.01910
Haiku 4.5 $0.00003 $0.00955

Measured 10d ago against content hash 616f69c23351, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade C, and why

terminal-use scanned grade C with 1 finding 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 10d 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.

Recursive force deletehighDestructive command

rm -rf with a variable or a broad path is one typo away from removing the wrong tree.

- `Allow command: rm -rf /?` → ASK USER
skills/terminal-use/SKILL.md · 886 lines

How it starts

The opening of the file, as written. The whole thing — 886 lines — stays where its author put it; the contents beside it link to each section on GitHub.

terminal-use: Terminal Control Skill

terminal-use-mcp v0.2.0 — This skill is maintained alongside the MCP server. Version check: call terminal.health → compare the version field with this header. If server version > skill version, prompt the user to update (npx skills update or re-download from GitHub). Full versioning details in the terminal-use-setup skill.

Terminal computer use — control interactive TUI programs that require keyboard input. This is NOT a shell runner. Use your bash tool for simple commands.

Customization Guide

This skill is designed to be trimmed to your needs. Each section is self-contained — delete any sections you don't need to reduce token consumption.

Section Content Safe to Remove?
Pre-flight Check Version/provider check before first use ⚠️ Keep — prevents errors from server mismatch or outdated skill
§1 What This Tool Is For Purpose, when to use/not use ⚠️ Keep — essential for correct tool usage
§2 Provider Selection native-pty vs tmux choice, provider availability ✅ Remove if you always use native-pty
§3 Standard Operation Loop Core workflow (snapshot → act → wait) ⚠️ Keep — critical for correct operation
§4 Tool Quick Reference Parameter tables for all available tools ✅ Remove if your AI already knows the tool schemas
§5 Available Key Names Key name list ✅ Remove — AI can call terminal.keys instead
§6 Safety Rules 6 non-negotiable safety rules ⚠️ Keep — prevents credential exposure and data loss
§7 Common Patterns 7 usage patterns with examples ✅ Remove largest section (~130 lines) if your AI learns by doing
§8 Snapshot Structure Response JSON fields ✅ Remove — AI can discover from tool output
§9 Error Handling Error code reference table ✅ Remove — AI can handle errors reactively
§10 Configuration & Setup Reference to terminal-use-setup skill ✅ Remove if terminal-use-setup skill is installed
§11 Checklist Pre-use decision checklist ✅ Remove if your AI follows §3 loop reliably
§12 Remote Terminal Control When to use remote, ACP vs SSH, provider selection ✅ Remove if you only use local terminals
§13 Remote Safety Rules 10 non-negotiable remote safety rules ✅ Remove if you only use local terminals (but ⚠️ keep if using remote)
§14 Remote Tools Reference targets / target_info / verify_target details ✅ Remove — AI can discover from tool schemas
§15 Remote Error Codes SSH-specific error code table ✅ Remove — AI can handle errors reactively
§16 Remote Operation Patterns 5 remote usage patterns with examples ✅ Remove largest section (~150 lines) if your AI learns by doing
§17 Remote Configuration Reference to terminal-use-setup skill ✅ Remove if terminal-use-setup skill is installed

Read the full file on GitHub · 886 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.

  1. 10d ago First seen · 886 lines · 26 tokens per session scan C 616f69c23351

Subscribe to this mod's changes

terminal-use is a skill published in the GitHub repository HLH2023/terminal-use-mcp (0 stars, last pushed 2mo ago), licensed MIT. It adds 26 tokens to every session and 9,550 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it C with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories

axstream

Fast deterministic macOS UI automation via replayable macros. Use BEFORE driving any macOS app's UI by hand (clicking/typing via computer-use or screenshots) — if a macro exists, replay finishes the task in seconds with no per-step reasoning; after doing a UI task manually, save it as a macro so the next run is fast.…

milind-soni/axstream · 96 tokens

thread-cron

Design and operate native thread-local cron jobs for a cron-enabled Codex CLI build.

PinkPig97/codex-thread-cron · 21 tokens

channel_message

Use this skill to proactively send a one-way message to a user/session/channel, usually only when the user explicitly asks to send to a channel/session or when proactive notification is needed. First query sessions with qwenpaw chats list, then push with qwenpaw channels send.

agentscope-ai/QwenPaw · 58 tokens

tmux

Remote-control tmux sessions for interactive CLIs by sending keystrokes, capturing pane output, and managing terminal multiplexer windows. Enables parallel coding-agent orchestration, background process management, and REPL interaction via sockets. Use when the agent needs to launch, monitor, or coordinate…

elizaOS/eliza · 87 tokens

things-mac

Manage Things 3 via the things CLI on macOS (add/update projects+todos via URL scheme; read/search/list from the local Things database). Use when a user asks Otto to add a task to Things, list inbox/today/upcoming, search tasks, or inspect projects/areas/tags.

elizaOS/eliza · 67 tokens

obsidian

Work with Obsidian vaults (plain Markdown notes) and automate via obsidian-cli. Use when the user asks about notes, vault management, PKM, knowledge base organization, wikilinks, or personal knowledge management in Obsidian.

elizaOS/eliza · 53 tokens