Montana Code

Several tools have heavily overlapping purposes: ask_pipeworx and ask_pipeworx_beta are described as currently identical, ask_pipeworx_grounded routes the same way, and deep_research also covers multi-source questions. Polymarket tools similarly blur (polymarket_edges vs polymarket_arbitrage vs bet_research), and ai_visibility_check vs scan_competitor_ai_presence overlap. The detailed descriptions help, but an agent would frequently struggle to pick the right one.

Almost all names use snake_case, but there is no consistent verb_noun pattern: some are verbs (ask_pipeworx, remember, forget), some are noun phrases (entity_profile, company_facts, mt_code_section), and some are mixed (polymarket_kalshi_spread, generate_llms_txt). The style is readable, but it does not form a predictable system.

37 tools is well beyond the typical well-scoped band, and many are meta-utilities (remember/recall/forget, subscribe/unsubscribe, pipeworx_feedback, pipeworx_trending, suggest_questions, discover_tools) that belong to a platform rather than a focused domain. The count feels heavy and diffuse rather than curated for a single purpose.

Given the server is named 'Montana Code,' the coverage is severely incomplete: a single mt_code_section lookup with no search, no titles list, no amendment history, and no other statutory features. The Pipeworx data side is fairly rich, but that only reinforces the mismatch—the stated domain of the server is largely unserved.