coding-agent

Delegate interactive coding work to Codex, Claude Code, or Pi agents with PTY support.

87|20|Updated Mar 9, 2026
One-click install
npx skills add https://github.com/openaeon/OpenAEON --skill coding-agent-openaeon
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: coding-agent
Source: https://github.com/openaeon/OpenAEON/tree/main/skills/coding-agent
Command: npx skills add https://github.com/openaeon/OpenAEON --skill coding-agent-openaeon

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Coding agents can be slow and disruptive without the right execution setup, causing broken output, hangs, and inefficient iteration. This Skill solves the problem by standardizing how to spawn and monitor interactive coding agents in the right directory with the required terminal support.

Core Features & Use Cases

  • PTY-first interactive execution: Ensures Codex, Claude Code, or Pi agents run correctly by allocating a pseudo-terminal (pty:true).
  • Background session orchestration: Starts long-running work in the background and monitors progress via process actions (poll/log), including interactive input handling (write/submit/send-keys/paste).
  • Safe workflow conventions: Uses scratch git repos for one-shots, supports PR review in temporary locations, and avoids unsafe workspaces (never in ~/clawd and never in ~/.openaeon or LIVE instances).

Quick Start

Start a one-shot build in a real project directory by running: delegate coding to an agent with a pseudo-terminal, using pty:true and workdir set to your project folder.

Frequently Asked Questions about coding-agent

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

FAQPage Schema
How do I run interactive coding agents like Claude Code or Codex without breaking terminal output?

To run interactive coding agents without broken output, delegate coding tasks using a pseudo-terminal with pty:true. This ensures agents like Codex and Claude Code maintain correct terminal behavior and reliable orchestration during execution.

What is the best way to orchestrate long-running background coding sessions for large refactoring?

The best way to orchestrate background coding sessions is starting agents in the background and monitoring progress via process actions. You can poll logs and submit interactive input like write, send-keys, or paste while handling large refactoring tasks.

Can I review pull requests in a temporary directory using coding agents?

Yes, you can review pull requests in temporary directories using coding agents. The workflow supports PR review in temporary locations and uses scratch git worktrees, avoiding unsafe workspaces like ~/clawd or ~/.openaeon.

Do I need git worktrees to safely delegate one-shot coding tasks to an agent?

You need git-aware scratch directories or worktrees to safely delegate one-shot coding tasks. This convention uses scratch git repos for isolated execution, ensuring your live instances remain untouched while the agent explores files iteratively.

Why does my coding agent hang or produce broken output when spawned in a project folder?

Coding agents hang or produce broken output when spawned without a pseudo-terminal. Allocating a pty ensures correct terminal behavior, allowing interactive agents to function properly without session disruptions or output formatting issues.