How do I install the Terminal skill?

Run npx skillmds@latest add majiayu000/terminal in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Terminal safe to use?

This skill has not completed SkillMD's automated safety review yet. Independent scanners report: SkillSpector: PASS, Skill Scanner: PASS. Capability flags: docs only. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Terminal?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Terminal free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Terminal?

majiayu000 (@majiayu000) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Terminal by majiayu000 — Agent Skill

Name: Terminal
Rating: 4.2 (5 reviews)
Author: majiayu000

Terminal

Execute shell commands in persistent, stateful VT100 terminal sessions. WHEN: User needs to "run a command", "execute shell", "build project", "run tests", "deploy", system administration. WHEN NOT: Simple file operations (use fs_* tools), reading known files (use fs_read_file).

majiayu000 Updated Sep 9, 2026 567 repo stars

File contents

terminal - Shell Command Execution

Core Concept

mcp__plugin_kg_kodegen__terminal executes shell commands in persistent VT100 terminal sessions. Terminals maintain environment variables, working directory, and shell state across commands. Use different terminal numbers for parallel work.

Actions

Action	Description	Required Parameters
`EXEC`	Execute command (default)	`command`
`READ`	Get current buffer snapshot	None
`LIST`	Show all active terminals	None
`KILL`	Gracefully shutdown terminal	None

Key Parameters

Parameter	Type	Default	Description
`action`	string	`"EXEC"`	Action to perform
`terminal`	number	0	Terminal instance (0, 1, 2...)
`command`	string	null	Command to execute (EXEC only)
`await_completion_ms`	number	300000	Max wait time (5 min default)
`clear`	boolean	true	Clear buffer before command
`tail`	number	2000	Max output lines to return

Usage Examples

Execute Command (default action)

{ "command": "ls -la" }

Execute in Specific Terminal

{ "terminal": 1, "command": "cargo build --release" }

Background Execution (fire-and-forget)

{
  "command": "npm run build",
  "await_completion_ms": 0
}

Execute with Timeout

{
  "command": "cargo test",
  "await_completion_ms": 60000
}

Read Current Buffer

{ "action": "READ", "terminal": 0 }

List All Terminals

{ "action": "LIST" }

Kill Terminal

{ "action": "KILL", "terminal": 0 }

Response Format

{
  "terminal": 0,
  "exit_code": 0,
  "cwd": "/project/path",
  "duration_ms": 1234,
  "completed": true
}

Parallel Work Pattern

Use different terminal numbers for concurrent tasks:

// Terminal 0: Build
{ "terminal": 0, "command": "cargo build" }

// Terminal 1: Tests (parallel)
{ "terminal": 1, "command": "cargo test" }

// Terminal 2: Watch logs
{ "terminal": 2, "command": "tail -f app.log" }

Background Execution

For long-running commands:

Fire-and-forget: await_completion_ms: 0
Check progress: { "action": "READ", "terminal": N }
Wait with timeout: If timeout occurs, command continues in background

Important Notes

Persistent sessions: Environment and working directory preserved
VT100 emulation: Actual rendered terminal output, not raw bytes
Automatic cleanup: Terminals cleaned up on connection close
State preservation: cd commands persist across calls

Remember

Terminals are numbered 0, 1, 2... - use different numbers for parallel work
Default action is EXEC - just provide command
Use await_completion_ms: 0 for background tasks
Use action: "READ" to check on background commands
Working directory persists - cd works across commands
For simple file operations, prefer fs_* tools (faster, safer)

---name: terminaldescription: Execute shell commands in persistent, stateful VT100 terminal sessions. WHEN: User needs to "run a command", "execute shell", "build project", "run tests", "deploy", system administration. WHEN NOT: Simple file operations (use fs_* tools), reading known files (use fs_read_file).--- # terminal - Shell Command Execution ## Core Concept `mcp__plugin_kg_kodegen__terminal` executes shell commands in persistent VT100 terminal sessions. Terminals maintain environment variables, working directory, and shell state across commands. Use different terminal numbers for parallel work. ## Actions | Action | Description | Required Parameters | |--------|-------------|---------------------| | `EXEC` | Execute command (default) | `command` | | `READ` | Get current buffer snapshot | None | | `LIST` | Show all active terminals | None | | `KILL` | Gracefully shutdown terminal | None | ## Key Parameters | Parameter | Type | Default | Description | |-----------|------|---------|-------------| | `action` | string | `"EXEC"` | Action to perform | | `terminal` | number | 0 | Terminal instance (0, 1, 2...) | | `command` | string | null | Command to execute (EXEC only) | | `await_completion_ms` | number | 300000 | Max wait time (5 min default) | | `clear` | boolean | true | Clear buffer before command | | `tail` | number | 2000 | Max output lines to return | ## Usage Examples ### Execute Command (default action) ```json { "command": "ls -la" } ``` ### Execute in Specific Terminal ```json { "terminal": 1, "command": "cargo build --release" } ``` ### Background Execution (fire-and-forget) ```json { "command": "npm run build", "await_completion_ms": 0 } ``` ### Execute with Timeout ```json { "command": "cargo test", "await_completion_ms": 60000 } ``` ### Read Current Buffer ```json { "action": "READ", "terminal": 0 } ``` ### List All Terminals ```json { "action": "LIST" } ``` ### Kill Terminal ```json { "action": "KILL", "terminal": 0 } ``` ## Response Format ```json { "terminal": 0, "exit_code": 0, "cwd": "/project/path", "duration_ms": 1234, "completed": true } ``` ## Parallel Work Pattern Use different terminal numbers for concurrent tasks: ```json // Terminal 0: Build { "terminal": 0, "command": "cargo build" } // Terminal 1: Tests (parallel) { "terminal": 1, "command": "cargo test" } // Terminal 2: Watch logs { "terminal": 2, "command": "tail -f app.log" } ``` ## Background Execution For long-running commands: 1. **Fire-and-forget**: `await_completion_ms: 0` 2. **Check progress**: `{ "action": "READ", "terminal": N }` 3. **Wait with timeout**: If timeout occurs, command continues in background ## Important Notes - **Persistent sessions**: Environment and working directory preserved - **VT100 emulation**: Actual rendered terminal output, not raw bytes - **Automatic cleanup**: Terminals cleaned up on connection close - **State preservation**: `cd` commands persist across calls ## Remember - Terminals are numbered 0, 1, 2... - use different numbers for parallel work - Default action is EXEC - just provide `command` - Use `await_completion_ms: 0` for background tasks - Use `action: "READ"` to check on background commands - Working directory persists - `cd` works across commands - For simple file operations, prefer `fs_*` tools (faster, safer)

Frequently asked questions