ucsahinn

Agentchef — Development skill for Claude Code

Development community

An unofficial, cross-platform setup kit for OpenAI Codex CLI and Anthropic Claude Code: specialist agents, curated skills, safe MCP defaults, preview-first installation, and clear verification.

How to install Agentchef

This entry records only its repository, not the path inside it, so there is no exact command to give. Open ucsahinn/agentchef and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Agentchef does

An unofficial, cross-platform setup kit for OpenAI Codex CLI and Anthropic Claude Code: specialist agents, curated skills, safe MCP defaults, preview-first installation, and clear verification.

Alternatives in Development

  • Native Feel Skill — An Agent Skill for designing cross-platform desktop apps that feel native — distilled from Raycast's 2.0 deep 1.9k ★
  • Creator Buddy — Creator Buddy: orchestrated Agent Skills for cross-platform content search, creator analysis, and viral trend 1.4k ★
  • Deep Research From Scratch — Deep research has broken out as one of the most popular agent applications 768 ★

README

AgentChef

AgentChef: a safer, clearer setup for terminal coding agents with agents, skills, MCPs, approvals, and verification

Validation workflow MIT license Two README languages Windows, macOS, Linux, and WSL

Release notes · Roadmap decision (ADR-006)

Read in: English · Türkçe

Getting a terminal coding agent running is the easy part. Turning it into a setup that stays clear, useful, and safe after the first week takes much more work.

I built **AgentChef** (formerly Codex Chef) because I kept solving the same setup problems: which agent should handle a task, which skill should guide it, which MCP is safe to use, what needs approval, and how to prove the result instead of trusting a confident answer.

AgentChef is an unofficial community starter: an open-source setup kit for two terminal agents, **OpenAI Codex CLI** and **Anthropic Claude Code**. It is built around the [official Codex documentation](https://developers.openai.com/codex) and the [official Claude Code documentation](https://code.claude.com/docs). It gives you a reviewed starting point without copying somebody else's private machine, credentials, sessions, or local memory.