Feeds

Most tools have clearly distinct purposes with detailed descriptions. The query variants (ask_pipeworx, ask_pipeworx_beta, ask_pipeworx_grounded) are well-differentiated, and each polymarket_* tool targets a specific analysis. Minor overlap exists (discover_tools vs suggest_questions, entity_profile vs compare_entities) but descriptions help.

All tools use consistent snake_case with recognizable patterns: verb_noun (ask_pipeworx, generate_llms_txt, resolve_entity), domain-prefix (polymarket_*, pipeworx_*), and noun phrases (entity_profile, recent_changes). No camelCase, mixed verbs, or cryptic abbreviations.

31 tools is well above the heavy threshold (25+). Many tools are unrelated to the server's 'Feeds' theme (memory tools, scan_dependency, generate_llms_txt, pipeworx_feedback, etc.), making the count feel bloated for the stated purpose. The name misrepresents the actual scope.

The data querying/research surface is very comprehensive: simple lookup, grounded answers, deep research, entity profiles, comparisons, recent changes, claim validation, entity resolution, and semantic search. The subscription lifecycle has subscribe/unsubscribe/list/recent_alerts, though it lacks update/pause and custom feed types, and the actual 5,521 data tools are hidden behind ask_pipeworx rather than exposed.