Openclaw Jobsearch

# OpenClaw Jobsearch `openclaw-jobsearch` is an OpenClaw skill for ranking live job listings from a single employer's careers portal against a job seeker's profile and stated preferences. The skill is designed for repeatable shortlist generation. Instead of relying on ad hoc prompting, it uses a fixed workflow to: - resolve an employer name through `EMPLOYERS.yaml` - inspect the employer's live careers portal - compare current roles against `USER.md` and `PREFERENCES.md` - return a ranked top-10 shortlist with evidence and source links ## Inputs The skill expects these files to exist at runtime: - `EMPLOYERS.yaml`: employer keys and careers portal URLs - `PREFERENCES.md`: the job seeker's preferences, constraints, and ranking priorities - `USER.md`: the job seeker's background, experience, and profile - `APPLICATIONS.yaml`: optional, used only when checking for duplicates or prior applications matters `USER.md` is assumed to be installation-managed by OpenClaw and may not live in this repository. ## Behavior The skill: - takes a single employer name - finds the matching careers portal in `EMPLOYERS.yaml` - reads live listings from that portal - filters roles using the preferences stated in `PREFERENCES.md` - scores roles using evidence from the job listing, `USER.md`, and `PREFERENCES.md` - returns the strongest matches first The scoring method and output contract are documented in: - [references/scoring-rubric.md](/Users/jordanmorris/projects/openclaw-jobsearch/skills/openclaw-jobsearch/references/scoring-rubric.md) - [references/output-format.md](/Users/jordanmorris/projects/openclaw-jobsearch/skills/openclaw-jobsearch/references/output-format.md) ## Invocation Invoke the skill as `openclaw-jobsearch`. Example: ```text Use openclaw-jobsearch to rank the top current jobs for OPENAI. ``` ## Output The skill returns: - a ranked shortlist of up to 10 roles - a fit score for each role - concise reasons each role matches - key risks or gaps - source links to the underlying postings - a short exclusions and limitations summary ## Limitations The skill depends on access to the employer's live careers portal. Results may be incomplete when: - listings are hidden behind heavy client-side rendering - the portal blocks browsing or search - work model or location details are missing from postings The skill does not invent missing facts. If a posting does not clearly state location, compensation, visa support, or similar details, the output should say so rather than guess.