Job Scout Skill

Instructions for the AI / OpenClaw when running this skill.

Purpose

• •Scout job postings and industry-related opportunities from configured sources.
• •Filter by keywords and relevance so only useful items are forwarded.
• •Summarize when needed so the morning digest is scannable.

Behavior

• •Use RSS (and optionally Playwright) via core.scraper to fetch job and industry news.
• •Use relevance (keywords + optional AI) via core.processor to keep only matching items.
• •Send the digest to Telegram using core.telegram and the pipeline’s chat_id.

Pipeline

• •Prefer pipeline config: pipelines/job_scout.json or pipelines/jobs.json.
• •If missing, the skill can fall back to defaults (sources and keywords) defined in run.py.

Output

• •One Telegram message per run: title, optional one-line summary, and link for each new item.
• •State is stored under state/seen_job_scout.json (or per-pipeline) to avoid duplicates.