A powerful orchestration skill that delegates coding tasks to specialized AI agents like Codex, Claude, and Pi via background bash processes.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install coding-delegate-agent
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 coding-delegate-agent 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 is a developer-centric orchestration layer designed to manage background AI processes for intensive programming tasks. By leveraging Openclaw Skills, it allows developers to offload feature creation, large-scale refactoring, and PR reviews to agents such as Claude Code, Codex, and OpenCode without blocking their primary workspace.
This skill excels at managing the lifecycle of these agents, utilizing pseudo-terminal (PTY) modes for interactive CLIs and sandboxed environments to ensure safety. It provides a robust framework for parallel execution, enabling a 'parallel army' of agents to handle multiple issues or PRs simultaneously via git worktrees and automated session monitoring.
Coding Agent Use Cases
- Building new features or entire applications from scratch using full-auto mode.
- Reviewing pull requests in temporary directories to maintain main branch integrity.
- Refactoring large, complex codebases across multiple files.
- Performing iterative coding tasks that require deep file system exploration.
- Executing parallel bug fixes across multiple git worktrees.
How Coding Agent Works
- The user initiates a coding task specifying the target agent (e.g., Codex or Claude) and a focused working directory.
- Openclaw Skills spawns a bash process, optionally enabling PTY mode for interactive terminal requirements.
- The agent runs the task in the background, returning a unique sessionId for tracking and monitoring.
- Developers monitor progress using process actions like log and poll to check status or provide manual input if prompted.
- Upon completion, the agent can trigger an immediate wake event to notify the user of the final results.
Coding Agent Setup
Ensure you have at least one supported agent binary installed (claude, codex, opencode, or pi). For Pi agent specifically:
npm install -g @mariozechner/pi-coding-agent
To run a basic one-shot task with Codex in a scratch directory:
SCRATCH=$(mktemp -d) && cd $SCRATCH && git init && codex exec "Your prompt here"
Coding Agent Data Schema & Taxonomy
The skill manages sessions through a process-oriented schema, allowing for detailed control over background tasks:
| Action | Purpose |
|---|---|
list |
Displays all active and recent background coding sessions. |
log |
Retrieves stdout/stderr output with offset support for real-time monitoring. |
poll |
Returns the current execution status of a specific sessionId. |
write/submit |
Handles stdin interaction for agents requiring user confirmation. |
workdir |
Restricts agent context to a specific path to prevent unauthorized file access. |
Coding Agent Advanced Features
- Multi-agent background support for running parallel PR reviews.
- PTY (Pseudo-Terminal) allocation for full compatibility with interactive CLI tools.
- Automated wake triggers using the system event tool to notify users immediately upon task completion.
- Git worktree integration for isolated, simultaneous issue resolution.
- Permission bypass modes for Claude Code to enable non-interactive automated workflows.
SKILL.md
Loading
Related Openclaw Skills
Powerpoint / PPTX
A sophisticated tool for programmatically creating, inspecting, and editing Microsoft PowerPoint presentations with a focus on template fidelity and visual quality.
Dashboard Setup Skill
A comprehensive framework and SOP for establishing business-ready dashboards with clear metrics and compliance standards.
Points Program Operations
A comprehensive operational guide and SOP for designing, implementing, and optimizing customer loyalty points programs.
Competitor Analysis SOP
A professional framework and Standard Operating Procedure for conducting actionable competitor analysis and market research.
BestYou Coach Widgets
A visualization-focused skill that renders BestYou health data into professional Dark Glass dashboards using the OpenClaw canvas.
Stablecoin Depeg Scanner
A rapid-response crisis evaluator for stablecoin depeg events, helping users identify asymmetric arbitrage opportunities based on collateral health and exploit analysis.
Healthcheck
A comprehensive system security hardening and risk-tolerance auditing tool for OpenClaw deployments.
Featured*
