resume-intelligence — Claude Code skill: resumes and interview prep

What resume-intelligence does

The skill runs the whole path from source inventory to finished files. It first writes source-inventory.md and tool-usage-log.md, then evidence.md, so you can see what was actually read and with which tools. On top of that it builds a resume, a detailed interview resume, a career master run, an ATS scorecard, a cover letter, LinkedIn recommendations and an interview prep pack.

A set of references loads on demand: references/source-model.md before pulling from more than one source, references/privacy-sanitization.md before writing final text from enterprise sources, references/enterprise-redaction-audit.md before finalizing anything from GitHub Enterprise, private Confluence or Jira. For local document folders it walks the approved root recursively with rg --files <folder>, falling back to find when rg is unavailable. It does not stop at the first subfolder.

The guardrails are explicit. It will not scrape LinkedIn behind a login or bypass platform controls, so bring a profile export or pasted text instead. Credentials stay out of plugin files, templates and evidence logs. Profile pictures and private documents are not uploaded to Canva or Figma automatically. Private repositories, Confluence and Jira count as confidential by default, and the skill asks before internal project names, Jira IDs or exact metrics go into user-facing output. If your resume draws on work projects, that is a reason to check what lands in the final file.