claude-code-runner_skill

This skill executes Claude Code tasks in non-interactive environments using PTY, auto-responds to prompts, and synchronizes project changes.

Python

2.5k

GitHub Stars

4

Bundled Files

2 months ago

Catalog Refreshed

3 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill openclaw/skills --skill claude-code-runner

_meta.json290 B
clawhub.json461 B
README.md3.0 KB
SKILL.md5.3 KB

Overview

This skill executes Claude Code programmatically in non-interactive environments using a PTY-based wrapper. It copies a project to a temporary workspace, runs Claude Code as a specified user, auto-responds to common interactive prompts, and syncs changes back to the original directory. It returns captured stdout/stderr and supports configurable timeouts and cleanup.

How this skill works

The skill creates a temporary copy of the provided working directory, adjusts ownership to a target user, and launches Claude Code inside a pseudo-terminal (PTY) so commands that require a TTY can run in containers and CI systems. It monitors process output, sends automated answers to known confirmation prompts, captures combined stdout/stderr, enforces a timeout, and synchronizes any file changes back to the source directory before cleaning up.

When to use it

Automated code review or refactoring in CI or batch jobs
Running Claude Code in containers or servers without an interactive terminal
Applying code changes programmatically while preserving file ownership
Tasks that may prompt for confirmations and need automated replies
Long-running code modification tasks where timeout control and cleanup are required

Best practices

Run from a Unix-like environment (Linux or macOS) with Python 3.8+
Ensure Claude Code is installed and available in PATH before invoking the skill
Provide a non-root user or run with sufficient privileges to change ownership and switch users
Increase timeout for large repositories or long analysis tasks to avoid premature termination
Extend the auto-response patterns for any custom interactive prompts your workflow produces

Example use cases

Automated code review of a microservice repository and returning a detailed log of findings
Refactor a legacy module (for example, convert raw SQL to an ORM) inside a CI job and commit the synchronized changes
Add or modify REST endpoints programmatically, including tests, then sync results back to the source tree
Run targeted bug fixes (memory leaks, race conditions) where the assistant must run tests and edit files
Run security scanning and automatic remediation scripts driven by a natural-language prompt

FAQ

It returns a string containing the combined stdout and stderr captured from the Claude Code execution.

What permissions are required to run this skill?

You need sufficient privileges to create temporary directories, change file ownership, and switch to the target user (root or sudo is typically required).