Snov

Most tools have clearly distinct purposes, backed by unusually detailed descriptions that explicitly compare them to related tools ('prefer over X'). However, there are overlapping families—three ask_pipeworx variants, multiple polymarket analysis tools, and several company-data tools—that could cause misselection if an agent skims descriptions. Overall, the descriptions rescue it from ambiguity, but it is not perfectly distinct.

All names use lowercase snake_case, which is internally consistent, but the verb-first vs noun-first pattern is mixed: compare_entities, discover_tools, validate_claim sit alongside company_facts, polymarket_edges, resolution_audit. The ask_pipeworx prefix provides a clear subfamily, and snov_ tools are grouped, but the overall set lacks a single predictable pattern. It is readable but not systematic.

With 39 tools, this exceeds the 25+ threshold for a heavy toolset. The server mixes Snov email lookup, Pipeworx data research, prediction-market analytics, memory, subscriptions, and meta-tools (discover_tools, suggest_questions, pipeworx_feedback), making it feel like a bundled multi-purpose kit rather than a focused single-domain server. Many tools could be consolidated or removed without losing core functionality.

For the broad scope it claims, the coverage is impressive: query/grounded research, company facts, entity resolution, comparison, claim verification, prediction-market analysis, memory, subscriptions, and Snov email lookups all have working lifecycle support. Minor gaps exist (no direct bet placement, no email campaign management, no general web fetch beyond what ask_pipeworx routes), but agents can work around these. No severe dead ends for the stated purpose.