shannon-thinking

What is shannon-thinking?

shannon-thinking is an MCP server that implements Claude Shannon's systematic problem-solving methodology. It provides a tool that helps break down complex problems into structured thoughts following Shannon's approach of problem definition, mathematical modeling, and practical implementation.

How to use shannon-thinking?

Install via NPX by adding the server configuration to your MCP client's JSON config. Invoke the single shannonthinking tool with a structured thought object specifying content, type, number, total thoughts, confidence level, dependencies, and assumptions.

Key features of shannon-thinking

• Iterative problem-solving with revision and recheck capabilities
• Dependency tracking between successive thoughts
• Assumption management with explicit documentation
• Confidence levels quantifying uncertainty per step
• Flexible validation combining formal proofs and experimental testing

Use cases of shannon-thinking

• Complex system analysis requiring structured decomposition
• Information processing problems needing theoretical frameworks
• Engineering design challenges with iterative refinement
• Optimization problems requiring both theory and practical implementation
• Cases where experimental validation complements formal proofs

FAQ from shannon-thinking

What methodology does the server implement?

It implements Claude Shannon's five-stage approach: problem definition, constraints identification, mathematical modeling, proof/validation, and implementation/experiment.

What tool does the server provide?

A single tool named shannonthinking that accepts structured thought objects with required fields like type, number, dependencies, assumptions, and confidence level.

What thought types are supported?

Five types: problem_definition, constraints, model, proof, and implementation.

What optional fields does the tool support?

It supports revision fields (isRevision, revisesThought), recheck definitions, proof elements, experimental validation with results and confidence, and implementation notes for practical constraints and solutions.

How can experimental validation be used?

The tool accepts experimentalElements with test description, results, confidence score (0–1), and limitations alongside formal proof elements.