chefcohen/corroborate-mcp

Overview

An MCP server that provides honest claim corroboration for AI agents. It counts independent sources, assigns confidence scores, and reports published error rates. It is keyless, read-only, and open source.

Best for

Best for
AI agents that require independent source counting and confidence scoring

Use cases

• Verify claims by counting independent sources
• Assess confidence scores for AI-generated statements
• Audit source diversity and error rates in research

Notes

An MCP server that provides honest claim corroboration for AI agents. It counts independent sources, assigns confidence scores, and reports published error rates. It is keyless, read-only, and open source.

1 stars on GitHub. Last updated 2026-07-29. Licensed MIT.

Use cases

• Verify claims by counting independent sources
• Assess confidence scores for AI-generated statements
• Audit source diversity and error rates in research

Pros

• Keyless operation simplifies setup
• Read-only design minimizes risk
• Open source enables transparency and customization

Cons

• Very early stage with minimal community traction
• Limited feature set based on current description
• No information on performance or scalability

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.