No user-submitted related servers found.
Scored across 38 tools
The CRUD and relation tools are clearly distinct (add/get/patch/delete/rename/merge), but the query surface has meaningful overlap: find_path, find_neighbors, find_backlinks, find_orphans, find_evidence, query_concepts, and query_ontology (which bundles 30+ graph operations) all serve discovery and could cause misselection. The code-analysis quartet (analyze_repo_structure, infer_imports, inspect_architecture, index_project) also has blurry boundaries, though each has a stated distinct role.
Naming is highly consistent: verb_noun snake_case throughout (add_concept, get_concept, list_concepts, patch_concept, merge_concepts). Batch variants consistently pluralize the noun (add_concepts, get_concepts, add_relations), and prefix pairs like connect_project_source/disconnect_project_source and find_path/find_backlinks/find_neighbors follow a clear predictable pattern.
At 38 tools this is well above the comfortable 3-15 range and over the 25+ heavy threshold, but the server's scope is genuinely broad: ontology CRUD, batch operations, relation management, git integration, code analysis, validation, and graph queries. Notably, query_ontology bundles ~30 operations into one tool, so the effective surface is even larger; the breadth earns some leeway but it remains a heavy, sprawling surface.
The tool surface covers the full ontology lifecycle: create (single/batch), read, update (patch/rename/reclassify), delete, merge; full relation management (add/remove/replace, batch); validation (validate_vault, validate_wiki); git checkpointing; a code-analysis pipeline; and rich graph-query operations. Minor gaps exist (no batch delete, no direct node export beyond compile_ontology), but there are no dead ends in the core workflows.