# Run CLI Commands in New Terminal Windows

Long-running CLI commands can occupy the active session and hide progress. This skill opens a separate terminal after checking command help.

## Install

```bash
npx skillstore add consiliency/spawn-terminal
```

## Metadata

- Status: approved
- Slug: consiliency-spawn-terminal
- Skillstore revision: r1
- Version status: missing
- Tree hash: 1fbe407c2fc07702fdf9c903bf83e63f48e1db312c6ff02d24f1a1a352cea9e1
- Author: Consiliency
- GitHub username: Consiliency
- License: MIT
- Repository: https://github.com/Consiliency/treesitter-chunker/tree/main/.ai-dev-kit/skills/spawn/terminal
- Ref: 635f69fb8d2f4e6330ba47a4e5a0fb239c04d110
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: confirmation\_required
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands, filesystem
- Quality score: 38
- Quality tier: warning
- Public page: https://skillstore.pages.dev/skills/consiliency-spawn-terminal
- Manifest: https://skillstore.pages.dev/api/skills/consiliency-spawn-terminal/manifest

## Capabilities

- Identifies the non-AI CLI command requested by the user.
- Runs the selected command with its help option before launch.
- Launches command strings in a separate terminal through fork\_terminal\(\).
- Supports blocking output capture for commands that must complete before returning.
- Can write complete terminal output to a debug file.

## Use Cases

- Convert Media in a Separate Terminal: Launch an ffmpeg conversion separately while preserving complete output in a debug file.
- Run Frontend Build Tasks: Start an npm build in another terminal without occupying the active assistant session.
- Capture CLI Diagnostics: Run a diagnostic command in blocking mode and return its captured output for review.

## Prompt Templates

### Launch a Basic Command

```
Open a new terminal and run [command]. Check its help first, then launch it without output capture.
```

### Save Terminal Output

```
Run [command] in a new terminal. Review its help first and save the complete output to a debug file.
```

### Wait for Captured Results

```
Start [command] in blocking capture mode after checking its help. Return the captured output and log file path.
```

### Prepare a Complex CLI Task

```
Assess this non-AI CLI task: [task]. Select a command, inspect help, explain important options, then launch it with requested capture settings.
```

## Limitations

- Depends on a companion fork\_terminal\(\) implementation outside this skill directory.
- Does not validate command safety beyond reviewing help output.
- Handles non-AI CLI processes and redirects AI agent requests to another skill.
- Opening a new terminal window depends on the host environment.

## Best Practices

- Review the exact executable, arguments, files, and destinations before launch.
- Enable file logging when command output is needed for later diagnosis.
- Use blocking capture only when the next task depends on command completion.

## Anti Patterns

- Do not launch destructive or privileged commands without explicit user confirmation.
- Do not treat help output as proof that a command is safe.
- Do not use this skill to start Claude, Codex, or another AI agent.

## Security Audit

- Audited at: 2026-07-19T04:19:43.216\+00:00
- Summary: All nine Ruby or shell backtick findings are false positives caused by Markdown formatting. Both path traversal findings are benign relative documentation references. However, the skill forwards user-selected commands without documented restrictions and can persist full command output.

## Stats

- Views: 238
- Downloads: 7
- Favorites: 0
- Popularity score: 0