AI Job Hunter

Problem: Low-Signal Job Searching

When resumes are scattered across platforms, job searches rely on batch keywords, and interview failures only lead to guessing, the process becomes noisy. AI Job Hunter turns it into a trackable loop: build a candidate profile, derive search terms, gather multi-source roles, score and filter them, then use funnel data to identify whether the issue is search direction, resume fit, or interview performance.

How It Works: From Profiling to Application Review

• Profile first: read MEMORY.md or prompt for resume, city, salary floor, and values, then apply Flower Chart, Iceberg, and Odyssey framing to set constraints.
• Multi-channel sourcing: combine web_fetch, web_search, and scripts for Liepin, Zhaopin, BOSS, Xiaohongshu, Zhihu, Maimai, government, and university channels, deduplicating identical company-role pairs.
• Quantified filtering: hard exclusions remove mismatches, while scores across skills, domain, front-line load, and salary produce S/A/B/C/D grades and role cards.
• Application and iteration: generate ATS keyword notes, resume variants, and outreach lines for high-fit roles, log confirmed applications, and review weekly funnel metrics.

Boundaries

It fits engineers or career switchers who need ongoing job-search tracking, especially when roles span job boards, social platforms, and public listings. It requires script dependencies; BOSS scraping needs Chrome and a first login, and some platforms may hide salary or full text. Confirm that automation is allowed by the target platform, and keep user login, job verification, and final submission under user control.