#533 · Primary category: AI Coding Assistants
qiaomu-goal-meta-skill
Turn vague or complex Codex tasks into strong `/goal` commands with outcome, verification, constraints, boundaries, iteration policy, completion evide
Project last updated:06/11/26
GitHub Stars
862
Forks
63
Contributors
1
License
MIT
Why we included this project
Anyone who uses Codex knows the pain of starting with a vague instruction and watching the agent drift. This skill installs into your agent setup and turns a loose request like "make me an app" or "fix this" into a concrete /goal command before work begins. The generated command covers what agents usually guess at on their own: how to verify the result, what not to touch, how many iteration rounds to allow, and when to pause and ask instead of pushing on. Rather than forcing you to fill out a template, it hands you a ready-to-copy recommended version with sensible defaults, plus a few multiple-choice tweaks for project shape, style, and scope. If you run many coding-agent sessions or manage someone else's Codex work, installing this and trying it on a task that used to take several retries is a good use of a few minutes.
Articles for this project
No articles for this project yet.
To suggest a topic or contribute an article, contact us.
Related projects in this category
superpowers
An agentic skills framework & software development methodology that works.
ECC
The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
skills
Skills for Real Engineers. Straight from my .agents directory.
opencode
The open source coding agent.
andrej-karpathy-skills
A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.