What problem it solves

People who want to monitor any public data source (jobs, prices, news, GitHub repos, sports) have to check manually or pay for scraping services. This skill guides Claude to build a fully automated agent that does it for free.

Who uses this workflow

Anyone in active job search, price hunting, competitive monitoring, or research who wants a personalised, automated feed without paying for hosting or LLM APIs.

What the skill does

• Designs the full project structure (scraper → AI enrichment → Notion storage)
• Writes scraper code for REST APIs, HTML pages, RSS feeds, and paginated sources
• Builds a Gemini Flash AI pipeline that batches 5 items per API call (stays within free tier)
• Adds a feedback learning loop — reads Applied/Rejected decisions from Notion and biases future scoring
• Deploys on GitHub Actions cron (free for public repos, zero infra)
• Auto-falls back across 4 Gemini models on quota exhaustion

Attribution / Inspiration

Built and battle-tested by @imrobinsingh as a job hunt agent for Chief of Staff roles.

Reference implementation: imrobinsingh/job-hunt-agent — scrapes 4+ job boards, batches Gemini calls, learns from Applied/Rejected in Notion. Fully working and free.

Example of how it's used

Build an agent that scrapes Chief of Staff job listings from LinkedIn and Cutshort,
scores each against my resume, and pushes results to Notion every 3 hours.

Result: A working Python project that runs automatically on GitHub Actions, with every Notion row containing AI Score, Match Score, Summary, and a personalised "Why I'm a Fit" note.