local-task-runner
by d-wwei
Coding Agents & IDEsBrief
This skill provides a mechanism to execute Node.js code snippets or full scripts locally on the host machine.
Description
What This Skill Does
Runs Node.js code snippets or full scripts directly on the host machine without spawning a subagent. Handles file management, cleanup, and timeouts automatically. Acts as the default execution method when subagent spawning is unavailable or inefficient.
Eliminates the overhead of manually writing, executing, and deleting temp files or spinning up a full subagent for simple code execution tasks.
When to Use It
- Running a quick calculation without launching a full subagent
- Checking local system status from within an agent workflow
- Executing a one-off utility script with automatic cleanup
- Enforcing a timeout on potentially long-running code
- Testing a Node.js snippet inline during agent task execution
Example Workflow
User asks: calculate the SHA-256 hash of a string without spawning a subagent
- 1
Construct Node.js code string using the crypto module to hash the input
- 2
Call run_task with the code string and an optional timeout value
- 3
Task runner writes and executes the code in an isolated context
- 4
Runner captures stdout and returns the result with task ID and duration
[TASK: abc123] Completed in 18ms --- STDOUT --- e3b0c44298fc1c149afb...
Installation
clawhub install d-wwei/local-task-runnerRun the CLI command, or paste the prompt into Claude Code, Cursor, Codex or whichever coding agent you use.