What is ai-product-project-coach?

This repository is an AI assistant skill designed to help people who are applying for AI product manager roles, switching careers into AI, or trying to make their personal projects look more competitive on a resume. When you interact with it, the assistant first figures out what you actually need. You might upload a resume, describe an existing side project, or just explain what you are confused about. From there, it routes your request to one of several types of guidance: suggesting a concrete AI project direction, helping you iterate on something you have already built, diagnosing your resume, answering common AI product manager interview topics, or producing a detailed plan for a complete AI project from scratch. The scope of projects it can help with is broad. It can review personal demos, web tools, retrieval-augmented generation (RAG) pipelines, agent workflows, and prompt-chaining setups. It can also help you prepare for the specific follow-up questions interviewers tend to ask about your projects, including how you would explain failures or edge cases. Internally, the skill is built as a set of reference documents and configuration files. One file routes your request to the right coaching mode. Others define the framework for assessing project quality, anonymized example cases, output templates, and the principles the assistant uses when judging what advice to give. The actual AI model is configured through an OpenAI YAML file. A notable design rule is that the assistant does not invent results. It will not claim your project went live or produced specific metrics unless your own materials say so. The goal is advice that holds up under a real interview. The repository is written in Chinese and is licensed under MIT.