|   shippedMCPTypeScriptNode.jsAutomation

Job Search MCP

A Model Context Protocol server that finds, scores, and drafts for real job postings.

A Google Sheet titled Job Search listing companies, roles, locations, work arrangement, category, a 1 to 100 fit score, application status, and posting URLs, sorted by fit score.
Ranked output: each role scored for fit and proximity, with the freshest, closest, best-matched postings on top.

Highlights

  • Four assistant-callable tools: find_jobs, score_job, rank_jobs, draft_cover_letter.
  • Reads Greenhouse / Lever / Ashby JSON APIs, so no captchas or scraping.
  • Scores by fit and East Bay proximity, with remote as its own tier.
  • Weights freshness so stale listings sink; scores each role 1 to 100.
  • Drafts plain cover letters with no em dashes.

Overview

An MCP server to run my own job search from inside an AI assistant. It pulls live openings from target companies, scores each for fit and proximity, weighs how fresh the posting is, drafts a cover letter, and ranks everything into the sheet below. It hits public applicant-tracking APIs (Greenhouse, Lever, Ashby) directly, which sidesteps the captchas and brittle scraping that break most job scrapers.