Stops AIs inventing remembered rules: file-only handoffs, independent hashed review, a paid-click ledger, and a project folder that stores the owner's verbatim rules.
xuyewild/multi-ai-skills
- Latest recorded source activity
- SkillsMP catalog refreshed
- skills collected
- 4
- GitHub stars
- 0
- GitHub forks
- 0
Install with an AI assistant
Copy this prompt into the AI assistant you're using.
Follow https://skillsmp.com/skill-install/prompt.md to install Agent Skills from https://github.com/xuyewild/multi-ai-skills.Loading the repository README…
Posts
See all →Skills in this repository
Showing 4 of 4 collected skills.
Use when several AIs (Claude, Codex, ChatGPT, subagents) work on one project. It sets who executes, who checks, and who decides. It hands work over through files, runs an independent review before each handoff, picks the model by task difficulty, logs paid…
Use for long projects that span many conversations and several AIs. The project folder is the memory. File names are fixed, and progress is whatever the files show. Human corrections are written back by type, and rules the owner has set go into a rule ledger.…
多个 AI(Claude、Codex、ChatGPT、子 agent)一起做一个项目时用:定谁执行、谁检查、谁拍板,按文件交接、交接前独立复核、按难度选模型、付费生成先记账再点,要人定的事一次用 ABCD 选项问清。适合 AI 视频、图像、代码等要花钱、要返工、跨多个对话的长任务;单次问答不用。
Source text: Chinese
跨多个对话、多个 AI 的长项目用:把项目文件夹当记忆,固定文件名,进度以文件为准;人的纠正按类写回,人定过的规则记进规则台账;新对话按文件判断接到哪一步。适合连续剧、系列视频、长文档、多轮代码等要做很多天的项目;单次问答不用。
Source text: Chinese