Provider-based shared memory runtime for AI clients, currently shipping Mem0 and Local JSON providers.
AgentMemory exposes 6 admin tools with minimal schema documentation and weak descriptions. All tools lack input parameter descriptions despite having parameter definitions. Tool names follow verb_noun convention (get_, list_, update_, delete_, pin_) but descriptions are generic and under 50 characters. No output schemas are documented. The codebase shows HTTP transport with a custom handler, but tool definitions appear inferred from function signatures rather than explicitly registered with full MCP schemas. Critical gaps: zero parameter descriptions, no output documentation, no error handling guidance.
Retrieve administrative statistics about the memory system
Delete a specific memory record with administrative access
Retrieve a specific memory record by ID with administrative access
List all memory records with administrative access
Pin a memory record to mark it as important
Update a specific memory record with administrative access
No parameter descriptions. All 6 tools have input parameters (memory_id, updates) but lack descriptions explaining what each parameter controls or accepts. LLMs cannot infer parameter meaning from names alone.
No output schemas documented. Tools return data but no schema defines what fields are in the response, their types, or structure. Agents cannot plan downstream calls or extract required data.
Tool descriptions are generic and under 50 characters. 'Retrieve administrative statistics' and 'Retrieve a specific memory record' lack context for when to use each tool vs alternatives, prerequisites, or what data is returned.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 37 | 2026-07-28+ | v2 |
No error handling guidance. Tools lack descriptions of failure modes, recovery steps, or actionable error messages. An LLM receiving a 404 or permission error has no guidance on what to do next.
Destructive operations lack confirmation pattern. delete_admin_memory and update_admin_memory can modify or destroy data but have no dry-run, confirmation, or undo mechanism documented.