No user-submitted related servers found.
Scored across 14 tools
Most tools are cleanly separated by resource+action: search_jobs vs salary_benchmark vs check_applications, and the watch trio (watch_intent / check_watches / manage_watch) is clearly delineated by lifecycle stage. The one genuine soft spot is employer-level introspection, where get_company_info (aggregate posting trust signals) and check_employer (legitimacy checklist, which also reports in_index) overlap; the descriptions do distinguish them, but an agent asking 'tell me about this company' could reasonably pick either. read_board vs add_employer vs refresh_index are adjacent but their descriptions make the read-only / persist / re-crawl split explicit.
Almost everything follows a snake_case verb_noun pattern (search_jobs, read_board, get_company_info, check_employer, list_portals, add_employer, remove_employer, check_watches, manage_watch, authorize_application). Deviations are minor: salary_benchmark is a noun phrase with no verb, and watch_intent reads as a noun while actually creating a watch (the delete/update counterpart lives in manage_watch). Nothing is camelCase or chaotic.
14 tools for a job-search domain that spans searching, watching, employer management, legitimacy checking, salary data and application follow-up is well-scoped and each tool maps to a distinct user need. No tool looks like a thin wrapper or a redundant duplicate of another.
Coverage is broad: search, full watch lifecycle (create/list/update/delete/check), employer management (read/add/remove/refresh/info/portal directory), salary benchmarking, and application tracking with authorize. Gaps are minor: there is no dedicated get_posting/job_id lookup (it is folded into search_jobs and read_board's ats_job_id, which is awkward to discover), and no way to list or discover employers in the index besides guessing a name or reading list_portals' blind-spot directory. Deliberate exclusions (no résumé submission) are documented rather than dead ends.