MCP Logical Solver

Example Use Cases

• 1

  Convert natural language logical problems into Prover9-compatible FOL and handle special constructs like XOR automatically.

• 2

  Combine LLM reasoning with automated theorem proving (Prover9/Mace4) to produce formally verified conclusions and explanations.

• 3

  Run batch verification of logical claims and output structured JSON results including premises, conclusions, tool usage, and explanatory reasoning.

Description

This project integrates Large Language Models with the MCP-Logic server and Prover9/Mace4 to process, translate, and formally verify logical problems expressed in natural language and First-Order Logic (FOL). It automates FOL transformations (including XOR handling), runs formal proofs, and interprets Prover9 outputs into human-readable answers (True/False/Uncertain) with detailed explanations. The system supports batch processing, structured JSON outputs, and configurable model providers (OpenAI, Anthropic, Gemini, Ollama). It is useful for workflows that require both LLM-driven analysis and rigorous formal validation.