ejentum/ejentum-mcp

Overview

Ejentum MCP server exposes 8 cognitive operations through 4 harnesses: reasoning, code, anti-deception, and memory. It runs in dynamic and adaptive modes, letting developers integrate structured cognitive checks into their workflows via the Model Context Protocol.

Best for

Best for
Developers building MCP-compatible agents that need lightweight cognitive guardrails

Use cases

• Add anti-deception validation to agent outputs
• Inject structured reasoning steps into LLM pipelines
• Persist and retrieve memory across MCP sessions

How to use

Install

npx @modelcontextprotocol/inspector npm run dev

Tools exposed

• anti-deception
• adaptive-reasoning
• adaptive-code
• adaptive-anti-deception
• adaptive-memory
• EJENTUM_API_KEY
• EJENTUM_API_URL

Tested with

Claude Desktop, Claude Code, Cursor, Windsurf, Cline, Continue

Notes

Ejentum MCP server exposes 8 cognitive operations through 4 harnesses: reasoning, code, anti-deception, and memory. It runs in dynamic and adaptive modes, letting developers integrate structured cognitive checks into their workflows via the Model Context Protocol.

8 stars on GitHub. Last updated 2026-05-31. Licensed MIT.

Use cases

• Add anti-deception validation to agent outputs
• Inject structured reasoning steps into LLM pipelines
• Persist and retrieve memory across MCP sessions

Pros

• Small, focused set of cognitive operations
• Adaptive mode adjusts behavior without manual tuning
• Open source with JavaScript implementation

Cons

• Limited to Ejentum API ecosystem
• Only 8 operations may not cover complex needs
• Low community traction (8 stars)

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