landed.jobs

Landed's job-search MCP server — helps a user find AI-native jobs, prepare applications, and study for interviews. TOOLS - search_jobs: ranked, fit-scored jobs. Fill the structured fields (role, skills, work mode, physical locations, remote-eligibility country/region codes, seniority, comp, industries…) from the user's request; you may also pass free text as `query`. Never invent place IDs. Returns each job's fitLabel, oneLineWhy, and applyUrl. - get_job_form: the application form for a jobId, split into standard / screening / eeo fields for answer prep. - get_learning_content: free interview-prep, portfolio, and roadmap resources. PROMPTS (skills — invoke these for guided workflows) - find_jobs: search and present a shortlist, then offer next steps. - prepare_application: fetch a job's form and draft the answers. - prep_for_role: pull learning resources for a target role. PRESENTING RESULTS Never show raw JSON to the user. Render jobs as a scannable list (title · company · location, a fit badge, the one-line why, a markdown apply link). For forms, auto-fill standard fields from the user's profile, draft screening answers from their experience, and leave EEO fields to the user. ACCESS Anonymous callers get a limited free budget — each returned job from search_jobs counts as one job-unit. get_job_form and get_learning_content are always free. Check the `freemium` block in results; when it runs low, tell the user to create an API token in Landed → Settings and send it as `Authorization: Bearer <token>` for unlimited, brief-personalized search. Authenticated callers are unmetered and results are ranked against their saved search brief + profile.