applylisted

# /apply — Quickfire Application You are running a quickfire application workflow for the user. The user will provide a company name and role (and optionally a JD URL or pasted JD). ## Inputs - `$ARGUMENTS` contains: `<company> <role>` and optionally a URL or pasted JD ## Steps ### 1. Parse input Extract company name, role title, and any JD URL/text from $ARGUMENTS. ### 2. Get the JD - If a URL was provided: fetch it with `curl` (browser User-Agent). If blocked, use `firecrawl scrape`. - If JD text was pasted: use that directly. - If neither: search the web for the job listing. Try `<company> <role> job listing site:greenhouse.io OR site:lever.co OR site:ashbyprod.com OR site:jobs.workable.com` first. - Save the JD to `USER/active/<company-slug>/jd.md` ### 3. Research (parallel agents) Run 2 agents in parallel: 1. **Company research**: What they do, funding, team size, recent news, tech stack if known. Save to `USER/active/<company-slug>/research.md` 2. **Role analysis**: What the JD actually asks for (explicit + implicit), what the hiring manager likely cares about, what would make them stop reading ### 4. Tailor CV - Read `USER/cv-master.md` for the bullet bank - Read `USER/references/` for structural examples - Map every JD requirement to a CV bullet. Flag gaps. - Build a tailored CV at `USER/active/<company-slug>/cv.md` - Follow the rules in `.claude/rules/cv-writing.md` and `.claude/rules/ai-writing.md` - Present the JD→CV mapping table to the user before finalisi