One-click install
npx skills add https://github.com/storyclaw-official/storyclaw-assistant --skill coding-agent-storyclaw-official
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: coding-agent
Source: https://github.com/storyclaw-official/storyclaw-assistant/tree/main/skills/coding-agent
Command: npx skills add https://github.com/storyclaw-official/storyclaw-assistant --skill coding-agent-storyclaw-official

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill automates complex coding tasks, code reviews, and refactoring by delegating them to specialized AI agents like Codex, Claude Code, or Pi, freeing up developers for higher-level problem-solving.

Core Features & Use Cases

  • Automated Code Generation: Build new features or entire applications with AI assistance.
  • Code Review & Refactoring: Delegate PR reviews and large-scale codebase refactoring to AI agents.
  • Iterative Development: Use agents for tasks requiring file exploration and iterative coding.
  • Use Case: You need to refactor a large Python codebase. Spawn the coding-agent skill to handle the refactoring, specifying the target directory and desired outcome.

Quick Start

Use the coding-agent skill to refactor the authentication module in your project directory.

Frequently Asked Questions about coding-agent

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I delegate code review and refactoring tasks to an AI agent?

You can delegate code review and refactoring to an AI agent by spawning a background process that handles pull request reviews in temporary directories or refactors codebases iteratively, respecting directory isolation to prevent workspace contamination.

Does the coding-agent skill work with Codex, Claude Code, and Pi?

Yes, the coding-agent skill works with Codex, Claude Code, and Pi. It requires specific execution modes for each, using PTY for Codex and Pi, and the --print mode for Claude Code to run background coding tasks.

What is the best way to automate building new features with AI coding agents?

Automating new feature generation is best handled by delegating coding tasks to AI agents via background processes, allowing them to perform file exploration and iterative coding to build new features or entire applications autonomously.

Can I use AI agents to review pull requests without affecting my main workspace?

Yes, you can review pull requests without affecting your workspace by running AI agents in temporary directories. This enforces directory isolation rules, preventing any workspace contamination during the automated code review process.

What are the limitations of using background AI agents for iterative coding?

The main limitation is execution mode compatibility; background agents require specific modes like PTY for Codex or Pi and --print for Claude Code, and they must strictly adhere to directory isolation rules to prevent workspace contamination.

How do I refactor a large codebase using an AI coding assistant?

To refactor a large codebase with an AI assistant, you spawn a coding task in a background process, specifying the target directory and desired outcome so the agent can iteratively explore files and execute the refactoring.