MCP server enabling persistent memory for Claude through a local knowledge graph
Server provides 9 tools with complete JSON Schema input definitions and basic descriptions. However, descriptions are minimal (10-60 chars, well below the 194-char baseline), lack actionable context for LLM selection, and omit prerequisites or when-to-use guidance. Parameter descriptions exist but are terse. No output schemas documented. Error handling is absent, operations fail silently or with generic exceptions. Tool naming follows verb_noun convention (create_, delete_, search_) but lacks clarity on destructive vs. safe operations in descriptions. Batch operations are well-designed (arrays of entities/relations), but no pagination, limits, or result capping for read_graph. No tool annotations (destructiveHint, readOnlyHint) despite clear risk levels.
Add new observations to existing entities in the knowledge graph
Create multiple new entities in the knowledge graph
Create multiple new relations between entities in the knowledge graph. Relations should be in active voice
Delete multiple entities and their associated relations from the knowledge graph
Delete specific observations from entities in the knowledge graph
Delete multiple relations from the knowledge graph
Open and retrieve specific nodes by name from the knowledge graph
Descriptions are critically short (10-60 chars vs. 194-char baseline). 'Create multiple new entities in the knowledge graph' lacks WHEN to use, prerequisites, or what the LLM should expect. LLMs cannot reliably select tools without richer context.
No output schemas documented. Callers cannot know what fields to expect from read_graph, search_nodes, or open_nodes. This forces LLMs to guess at response structure and risks parsing errors.
No error handling or recovery guidance. If create_entities fails (e.g., duplicate name, invalid entityType), the tool throws a generic error with no actionable next step for the LLM. Missing pattern:recovery-guide.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 46 | 2026-07-28+ | v2 |
Read the entire knowledge graph
Search for nodes in the knowledge graph by name, type, or observations
read_graph has no pagination, limit, or result capping. A large knowledge graph could return thousands of entities/relations, exhausting context window. No mention of limits in description.
Destructive tools (delete_*) lack confirmation or dry-run support. An LLM could accidentally delete all entities with no undo path. Missing pattern:confirmation-request.