Chinese AI resume

Create a factual, single-page Chinese resume that helps an interviewer understand what the candidate built, why key choices were made, what failed or remained incomplete, and what can be verified. Default output is an editable DOCX plus a PDF; use the blue technology-style template only when the user wants this visual direction or gives no conflicting design reference.

Decide the task path

1. Create a resume from source material. Follow the full workflow below.
2. Revise an existing resume. Extract the current facts, compare them against repositories and supplied evidence, then replace weak or unsupported claims. Keep a user-approved visual reference unless the user requests a redesign.
3. Create a targeted variant. Keep the same facts. Change project order, title, summary, skills, and emphasis for the target role. Do not invent a tool, metric, company outcome, or production deployment to force a match.

Full workflow

1. Collect the fact base before drafting

Request or locate: the target job direction, contact details, education, experience, GitHub account, portfolio links, and user-approved project documents. If the candidate authorizes GitHub access, inspect the repositories before choosing flagship projects.

Create a fact sheet for each serious candidate project using references/project-fact-template.md. Record architecture, personal contribution, quality evidence, and known limitations separately. Treat README claims, test reports, source code, commits, demos, and CI output as evidence. Treat unverified descriptions as leads, not resume facts.

Rank projects by all of the following: relevance to the target role, technical depth, readable public evidence, test or demo quality, and ability to explain decisions. Choose at most three flagship projects. Put small experiments, skills, tutorials, or tangential work in “其他作品”.