← ClaudeAtlas

resume-checklisted

Score a resume against a job description (keyword match %, same idea as Jobscan), sort every missing keyword into "have it, add the word" / "ask first" / "don't have it", and offer to make an updated resume. Takes a JD (file path, URL or pasted text) and an optional resume path (.pdf or .html; defaults to resume/master.html). Never edits the master resume — an update always goes into a tailored copy. Use when the user says "/resume-check", "score my resume against this JD", "resume match score", or "how well does my resume match this job".
vikas9dev/claude-job-hunt · ★ 0 · Data & Documents · score 75
Install: claude install-skill vikas9dev/claude-job-hunt
# Resume Check Write the report in short, plain sentences with common words. Technical terms stay exact. ## Step 0 — Preconditions 1. Read **`.claude/skills/_shared/PROFILE_AND_FILTERS.md`**. If it does not exist, stop and say: `No profile yet. Run /setup-profile first (it copies the template and asks you a few questions).` 2. Read **[`_shared/RESUME_RULES.md`](../_shared/RESUME_RULES.md)**. It has the file paths, what the score means, the three buckets, the allowed edits, and how to get the JD. ## Input ``` /resume-check <jd> [<resume-path>] ``` - `<jd>`: a file path, a URL, or `paste` (then ask for the JD text and wait for it). If the message has JD text as well as a path, use the pasted text. - `<resume-path>`: `.pdf` or `.html`. **If it is missing, use `resume/master.html`** and say so in one line. If `resume/master.html` does not exist either, stop and tell the user to copy `resume/master.template.html` to `resume/master.html` and fill it in (or run `/setup-profile`). ## Steps ### 1. Get the JD and save it Follow *Getting the JD* in RESUME_RULES. Save the text to `<scratchpad>/resume-check/jd.txt`. From the JD, note the company, role title and seniority. ### 2. Score ```bash python3 scripts/resume_match.py score --jd <jd.txt> --resume "<resume-path>" --json ``` ### 3. Judge it yourself The script counts words. You add the judgment: 1. **Sort every missing term** into ✅ / 🟡 / ⛔ (RESUME_RULES → *Sorting each missing term*). Use `related_evidence`