Adaptive Graph of Thoughts
JSON →An intelligent scientific reasoning framework that uses graph structures and Neo4j to perform advanced reasoning via the Model Context Protocol (MCP).
Tools · 5
- process_query Process a scientific query through the ASR-GoT 8-stage reasoning pipeline, generating hypotheses, integrating evidence, and producing a structured answer stored in the Neo4j graph database.
- get_graph_state Retrieve the current state of the reasoning graph for a given query, including nodes, edges, and confidence scores.
- explore_graph Explore the knowledge graph interactively, allowing traversal of nodes and relationships to inspect reasoning paths and evidence.
- natural_language_query Submit a natural language query to the system, which is processed via the NLQ endpoint for flexible question answering.
- health_check Check the health status of the MCP server, including connectivity to Neo4j and external services.
Environment variables
OPENAI_API_KEYANTHROPIC_API_KEY
★ 30 GitHub stars