Browse / Productivity Workflow / Fork Terminal

Fork Terminal

Opens a new terminal window to run commands, delegate tasks to other AI agents, or start parallel workstreams.

SkillProductivity WorkflowTask ManagementOrchestrationCliAi Agent

The source repository doesn't declare a license. Check its terms before reusing the code.

Key features

  • Provides context handoff by summarizing the current conversation for new agentic sessions.
  • Parses natural language requests to identify the desired tool, command, and task.
  • Spawns independent, parallel terminal sessions for multitasking.
  • Uses a configurable, cookbook-based approach for constructing tool-specific commands.
  • Supports multiple agentic coding tools (Claude Code, Codex, Gemini) and raw CLI commands.

Use cases

  • Delegating a complex coding task (e.g., 'fix tests') to a new AI agent session while continuing the main conversation.
  • Starting a long-running process like a development server or data script in a separate window.
  • Running a quick CLI command (e.g., `curl` or `ffmpeg`) in a new terminal without interrupting the current workflow.

FAQ

How does this skill improve my workflow?

It significantly boosts productivity by enabling parallel workstreams. Its key feature is 'context handoff,' which automatically summarizes your current conversation for a new AI agent, so it understands the task without needing a full re-explanation.

When should I use this skill?

Use it when you need to multitask without disrupting your current workflow. It's ideal for running tests in a separate window, starting a background process, or assigning a sub-task like refactoring code to another AI agent while you continue your primary work.

What does the Fork Terminal skill do?

This skill opens a new, independent terminal window from your current Claude session. You can use it to run standard command-line tools or delegate complex tasks to other AI coding agents like Claude Code, Codex, and Gemini.

What capabilities does it provide?

It supports spawning terminals with raw CLI commands (e.g., `ffmpeg`, `curl`) and multiple AI agents. The skill parses natural language to identify the right tool and task, provides conversational history as context, and uses a flexible, cookbook-based system to construct commands.