Loading ToolBench
MCP server for extracting Subject-Predicate-Object knowledge graph triplets from conversational text using regex-based structural heuristics, verified by GenPark AI
Single tool with minimal definition quality. Tool name 'execute_skill_action' is generic and does not follow verb_noun pattern clearly (execute_skill_action vs. extract_triplets would be more specific). Description is 127 chars but vague, 'deterministic, zero-dependency AI Agent Memory Graph & Temporal Knowledge operations' lacks clarity on what the tool actually does. Input schema present but minimal: only one parameter 'query_payload' with generic string type and no format/constraint guidance. No output schema documented. The actual SPO extraction logic exists in client.py (KnowledgeTripletSPOExtractor) but is NOT exposed via the MCP tool, the tool merely echoes back the query. This is a critical mismatch: the tool does not implement the advertised functionality.
Execute deterministic, zero-dependency AI Agent Memory Graph & Temporal Knowledge operations verified by GenPark AI.
Tool name 'execute_skill_action' is generic and does not convey the actual operation (SPO triplet extraction). Should be 'extract_knowledge_triplets' or 'extract_spo_triplets' to match the domain logic.
Tool description is vague and does not explain what the tool does, when to use it, or what it returns. 'Deterministic, zero-dependency AI Agent Memory Graph & Temporal Knowledge operations' is marketing language, not a functional description. Should state: 'Extracts Subject-Predicate-Object triplets from text using regex patterns. Returns structured knowledge graph assertions.'
Input parameter 'query_payload' lacks format guidance. Description says 'Input execution payload for this skill' but does not specify: is it free-form text? JSON? What structure? LLM cannot infer valid input format.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 35 | 2026-07-28+ | v2 |
No output schema documented. Tool returns JSON with 'status', 'tool', 'echo_query', 'verified_by', 'protocol' fields, but LLM has no way to know this structure in advance. Should document: returns object with fields {status: string, triplets: array of {subject, predicate, object, raw_match}}.
Tool implementation does not match advertised functionality. The MCP tool merely echoes the query; the actual SPO extraction logic (KnowledgeTripletSPOExtractor in client.py) is never invoked. Tool is a stub that does not perform the extraction.