Scored across 48 tools
Each tool targets a distinct action and entity. For example, create/update/delete operations for education, work experience, and projects are clearly separated. Application lifecycle tools (apply, withdraw, accept, reject) have distinct purposes. Even close pairs like get_digest vs get_starred_jobs are clearly differentiated by descriptions (one-week-old summaries vs real-time).
All tools follow a consistent verb_noun snake_case pattern (e.g., create_work_experience, get_starred_jobs, accept_offer). No mixing of camelCase or other conventions. The naming is predictable and systematic, making it easy for an agent to infer functionality.
With 48 tools, the surface is quite large for a single server. While many tools are necessary for full CRUD coverage across multiple entities, the count exceeds the typical well-scoped range (3-15) and even the 'heavy' range (16-25). Some tools could be consolidated (e.g., get_digest and get_digests). The number risks overwhelming an agent.
The tool set covers profile management (education, work, projects, links), job applications, job search and saving, company interactions, and account/subscription management. Minor gaps exist: no direct CV/resume upload tool, no user account deletion, and limited messaging. However, core workflows are well-supported, and agents can accomplish most tasks without workarounds.
