No user-submitted related servers found.
Scored across 82 tools
Many clusters of near-duplicate tools: five record_* outcome tools, five verify_*/eicv_* verification tools, and eleven work_* tools with cryptic distinctions (e.g., work_record_context vs work_compile_context vs work_record_memory). Multiple retrieval searches (vault_search, vault_query, recall_relevant, entroly_retrieve, shared_memory_search) create significant confusion about which to call.
Mostly snake_case, but conventions are inconsistent: verb_noun tools (create_receipt, verify_beliefs) mix with brand-prefixed (entroly_retrieve, entroly_dashboard), noun phrases (blast_radius, smart_read), and a huge spread of prefixes (vault_, work_, shared_memory_, eicv_, record_, verify_, compile_, etc.). The pattern is not predictable across the tool surface.
82 tools is an extreme mismatch for a context engine. The server sprawls across context optimization, vault/belief management, hallucination verification, security scanning, work-graph bookkeeping, shared memory, shell/image compression, and more—far beyond what a coherent MCP server should expose.
The core context engine (optimize/recall/receipt) is decently covered, but the overall surface is a mishmash of unrelated domains with uneven lifecycle support—many artifacts can be created/queried but not updated or deleted, and several tools (SAST, image compression, shell compression) belong to entirely different problem spaces. Significant gaps exist for any single coherent stated purpose.