Scores unscored jobs using Qwen2.5:7b model and marks duplicates from the same company within 30 days.
npx skills add https://github.com/MBojer/OpenClaw_JobHunter --skill job-filter
Please help me install this Agent Skill. Skill: job-filter Source: https://github.com/MBojer/OpenClaw_JobHunter/tree/main/skills/job-filter Command: npx skills add https://github.com/MBojer/OpenClaw_JobHunter --skill job-filter
Score unscored jobs against the user's profile using a local Qwen2.5:7b model. After scoring, runs a dedup pass to detect cross-posted jobs from the same company and stores results in the database.
Run the scoring script to process new jobs and perform dedup.
High-intent search queries and answers about installing and using this skill.
You can score job postings locally by running the scoring script against a local Qwen2.5:7b model, which evaluates unscored jobs based on your user profile and writes the results directly to your database.
Deduplicating cross-posted jobs is handled by a dedup pass that identifies duplicate postings from the same company within the last 30 days and marks them accordingly in your database.
Yes, you need a local Qwen2.5:7b model deployed in your environment to process the job scoring, as the workflow relies entirely on local execution and prompt files for deterministic results.
Yes, the scoring workflow exposes configurable flags including limit, rescore, job-id, and skip-dedup, allowing you to target specific jobs for re-scoring and bypass the dedup pass if needed.
Prompt engineering drives deterministic scoring by using specific prompts from the skill's prompt files to guide the local Qwen2.5:7b model in evaluating unscored jobs against your profile.
Using a local 7b model for job scoring limits processing scale compared to larger models, but ensures data privacy by evaluating unscored jobs locally and storing results directly in your PostgreSQL database.
offlinecvScrape, rank, and import job postings into an offlinecv database.
CapatainaFetch job listings from ATS APIs and scrape company career pages.
Awannaphasch2016Design B2B landing pages using a 14-section psychological sequence.
CarbeneAIRoute sensitive research queries to a local Ollama API endpoint.
