Scored across 25 tools
Each tool targets a clearly distinct research domain or workflow stage: web, news, academic, legal, patent, local, biomedical, economic, and image search are all cleanly separated, and adjacent tools like verify_citation, audit_bibliography, and verify_recommendation are differentiated by scope. Cross-references between tools further reduce ambiguity, so an agent should rarely mis-select.
The dominant <domain>_search pattern is consistent across most search tools, and verification/formatting tools use predictable verb prefixes like verify_, audit_, and format_. A few names like paper_fulltext, citation_graph, research_export, and search_and_scrape deviate from the pattern, but all names remain readable and generally predictable.
25 tools is at the high end of what feels reasonable for a server, but the broad "web researcher" purpose justifies the breadth: each tool covers a distinct research domain or lifecycle step rather than duplicating another. It is slightly heavy, but the set is well-scoped enough that no tool feels redundant.
The tool surface covers the full research lifecycle: discovery across multiple verticals, full-text retrieval, session tracking and recovery, export, bibliography formatting, citation verification, archiving, and specialized audits. There are no obvious dead ends—sessions can be recovered, exported, and converted into verified, formatted citations.