coding-agent

Orchestrate background coding-agent sessions for interactive development workflows.

5|Updated Feb 8, 2026
One-click install
npx skills add https://github.com/tiancaiamao/ai --skill coding-agent-tiancaiamao
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: coding-agent
Source: https://github.com/tiancaiamao/ai/tree/main/skills/coding-agent
Command: npx skills add https://github.com/tiancaiamao/ai --skill coding-agent-tiancaiamao

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill removes the friction of running interactive coding assistants from a messaging-driven workflow by letting OpenClaw start, monitor, and control external coding agents for you.

Core Features & Use Cases

  • Background agent control: Launch Codex, Claude Code, Kiro CLI, OpenCode, or Pi in a managed session.
  • Multi-turn task handling: Keep long-running coding conversations alive, send follow-up prompts, and capture logs.
  • Real-world use case: Use it to refactor a module, review a pull request, or drive a project-specific coding task without leaving OpenClaw.

Quick Start

Ask OpenClaw to use this skill to start the coding agent you want in your project directory and keep the session running in a PTY-backed background process.

Frequently Asked Questions about coding-agent

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

FAQPage Schema
How do I run Claude Code or Codex in the background for long refactoring tasks?

You can run Claude Code, Codex, Kiro CLI, OpenCode, or Pi in PTY-backed background sessions managed by OpenClaw to keep long-running refactoring conversations alive and capture logs.

What is the best way to manage multi-turn coding agent sessions without leaving my messaging workflow?

Managing multi-turn coding agent sessions involves starting a PTY-backed process in a scoped workdir, sending follow-up prompts, and monitoring logs to keep interactive agents stable and controllable.

Can I scope a coding agent to a specific project directory and monitor its session?

Yes, coding agent sessions require workdir scoping and session monitoring to ensure interactive agents like Claude Code or Codex remain stable and controllable within your project directory.

How do I keep interactive coding agents stable during one-shot prompts or multi-turn debugging?

Interactive coding agents stay stable during one-shot prompts and multi-turn debugging by using PTY-backed terminal sessions and trust-tool flags to control the background process.

Do I need PTY-backed terminal sessions to run coding agents like Kiro CLI or Pi?

Yes, PTY-backed terminal sessions are required to run coding agents like Kiro CLI or Pi, ensuring the interactive background processes remain stable and controllable.

Why does my interactive coding agent session drop when running long-running refactors?

Long-running refactors require PTY-backed terminal sessions and trust-tool flags to prevent interactive coding agent sessions from dropping and to maintain controllability.