elowwwen/resume-evidence-workflow

A complete resume workflow for various agents and conversational AI, covering professional experience inventory, job description (JD) matching, resume rewriting, and submission version generation.

It first helps users discover easily overlooked experiences and values, building a reusable Markdown experience evidence base; After obtaining the target job description (JD), it interprets job requirements, matches evidence, and rewrites according to job needs; When specific writing methods or document generation capabilities are needed, other skills are introduced.

Main Features

• Even if you don’t have a job description (JD), you can first review your experience and don’t postpone all your digging until the application stage.
• Use natural, specific questions to help users recall, while internally retaining professional judgment.
• Distinguish between facts, estimates, interpretations, and resume presentation options, without supplementing experiences out of thin air.
• First, create a job description (JD)—an evidence matching matrix—then decide on the emphasis, compression, or rewriting direction of the resume.
• Executed uniformly by this workflow, borrowing specific methods or file generation capabilities from other Skills as needed.
• New facts, updated data, and project progress discovered during the rewriting phase are synchronized back to the experience database and explained to users.
• After the first resume analysis, you can save 基础简历.md and 简历版式档案.md; In the future, when processing new job descriptions (JDs), you don’t need to upload the same PDF repeatedly.
• During in-depth exploration, appropriate questions are asked about common interview topics such as collaboration, difficulties, decision-making, and review. Retention of experiences can also assist in future interview preparation.

How to use it

Method 1: Give the repository link to the agent

Copy the repository link and directly tell the Agent:

If the Agent supports installing Skills from GitHub, you can also clone or install the entire repository:

During installation, just keep the original directory structure of the warehouse.

Method 2: Upload to Conversational AI

After downloading the repository, upload it to the conversational AI:

• SKILL.md
• All Markdown files in the references folder
• If you need to generate editable HTML, upload it assets/editable-resume-base.html

Then tell it:

agents/openai.yaml is mainly used for Agents that support skill configuration; Ordinary conversational AI does not require uploading. README.md and LICENSE are also not required to run workflows.

After completing the first experience organization and resume import, you can save the AI-generated 经历库.md, 基础简历.md, and 简历版式档案.md files together. Next time you switch conversations or agents, just upload these three Markdown files; usually, you don’t need to re-upload the original PDF.

Reusable: Optional directional masters and project material libraries

When building your initial resume, the workflow will ask you which positions or directions you mainly apply for; Once explained, just reuse them. If you repeatedly apply for different directions, you can choose to create several parent directions, such as brand marketing, community operations, or user research. If you only have one direction, one is enough; the quantity depends on your goals.

The direction master page stores commonly used experiences, selection, order, and expression in that direction, reducing the need to reorganize from scratch and maintaining narrative coherence.