A powerful orchestration tool for delegating multi-file coding tasks and reviews to specialized AI agents in isolated environments.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install openclaw-coding-agent-playbook
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install openclaw-coding-agent-playbook using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Coding Agent?
The Coding Agent skill provides a robust framework for offloading intensive development work to specialized AI coding engines such as Codex, Claude Code, Pi, and OpenCode. As a core component of the Openclaw Skills ecosystem, it enables developers to manage complex code transformations while maintaining safety and process control.
This skill is designed to handle the heavy lifting of multi-file edits and large-scale refactors by providing a standardized interface for interacting with third-party coding agents. It supports various execution modes, including background processing and pseudo-terminal (PTY) environments, ensuring that tasks can run autonomously while remaining fully observable and interactive when necessary.
Coding Agent Use Cases
- Building complex features that require coordinated edits across multiple files.
- Executing large-scale refactors in parallel workstreams to improve code quality.
- Performing PR reviews and issue triage within isolated git worktrees or temporary clones.
- Implementing repeatable run, fix, and verify loops for continuous integration tasks.
How Coding Agent Works
- The user identifies a development task and selects the most suitable AI coding agent supported by Openclaw Skills.
- The skill configures the execution environment, determining if a pseudo-terminal (PTY) is required based on the agent's requirements.
- A session is initiated in the target repository or a temporary workspace to ensure the live environment remains stable.
- For long-running tasks, the agent runs in the background while the user monitors progress via real-time logs and status polls.
- Upon completion or when blocked, the skill facilitates feedback loops and finalizes the code changes for manual review.
Coding Agent Setup
To utilize this skill, ensure you have at least one supported agent installed (claude, codex, opencode, or pi). You can initiate tasks directly via the CLI:
# Basic command for Claude Code
bash workdir:~/project command:"claude --permission-mode bypassPermissions --print 'Refactor auth middleware'"
# Command for agents requiring PTY
bash pty:true workdir:~/project command:"codex exec --full-auto 'Add request timeout handling'"
Coding Agent Data Schema & Taxonomy
The Coding Agent organizes its operations around process sessions and repository state. The following table describes the data management structure:
| Component | Details |
|---|---|
| Session Management | Tracked via unique session IDs for background process lifecycle |
| Workspace | Isolated workdirs or git worktrees to prevent accidental data loss |
| Logging | Continuous log stream capture for auditing and troubleshooting |
| Metadata | Integrated with Openclaw Skills taxonomy for agent requirements |
Coding Agent Advanced Features
- Background session monitoring with the ability to submit manual input to blocked processes.
- Safe review workflows using temporary clones and git worktrees to isolate PR analysis.
- Standardized prompt templates for complex tasks including constraints and required verification checks.
- Failure recovery protocols such as session polling and explicit scope tightening for stalled agents.
- Support for Openclaw Skills system events to trigger notifications upon task completion.
SKILL.md
Loading
Related Openclaw Skills
Rebranding Strategy
A comprehensive guide and workflow for executing a seamless rebrand, from technical domain migrations to multi-channel communication strategies.
Hire YouTube Video Editor
A comprehensive hiring assistant designed to help YouTubers find, vet, and onboard professional video editors tailored to their specific style and budget.
Vheer AI Assistant
An intelligent AI companion designed to streamline content creation, strategy planning, and workflow automation on the Vheer platform.
LipSync Video AI
An AI-powered utility that automatically matches audio tracks to speaker lip movements in video files using high-performance cloud GPU processing.
Domain Selection
A technical guide for choosing SEO-optimized domains that balance brand identity with search engine performance.
Domain Architecture
A strategic framework for deciding between subfolders, subdomains, and independent domains to optimize brand architecture and SEO authority.
Featured*
