MCP server for the Ejentum Logic API. Exposes the four cognitive harnesses (reasoning, code, anti-deception, memory) as MCP tools any agentic client can call.

Letta integration for the Ejentum Reasoning Harness. Four self-contained Python functions (harness_reasoning, harness_code, harness_anti_deception, harness_memory) registered with a Letta server via tools.upsert_from_function.

Mastra integration for the Ejentum Reasoning Harness. createEjentumTools() returns four agent-callable tools (reasoning, code, anti-deception, memory) for Mastra Agent. Lighter alternative to the MCP route.

LlamaIndex tool spec for the Ejentum Reasoning Harness. Wraps the hosted Ejentum MCP server and exposes four cognitive harness tools (reasoning, code, anti-deception, memory). Each call returns a task-matched cognitive operation engineered in two layers: natural-language procedure plus executable reasoning topology.

Vercel AI SDK integration for the Ejentum Reasoning Harness. createEjentumTools() returns four agent-callable tools (reasoning, code, anti-deception, memory) for generateText/streamText. Lighter alternative to the MCP route.

PydanticAI toolset for the Ejentum Reasoning Harness. EjentumToolset subclasses FunctionToolset and registers four tools (reasoning, code, anti-deception, memory) the agent calls before generating. Each call returns a task-matched cognitive operation engineered in two layers: natural-language procedure plus executable reasoning topology.

Agno Toolkit for the Ejentum Reasoning Harness. Four cognitive harnesses (reasoning, code, anti-deception, memory) as agent-callable methods. Each call returns a task-matched cognitive operation engineered in two layers: natural-language procedure plus executable reasoning topology.

LangChain integration for the Ejentum Reasoning Harness. Four BaseTool subclasses (reasoning, code, anti-deception, memory) plus an EjentumTools factory. Each call returns a task-matched cognitive operation engineered in two layers: natural-language procedure plus executable reasoning topology.

Genkit integration for the Ejentum Reasoning Harness. createEjentumTools(ai) registers four agent-callable tools (reasoning, code, anti-deception, memory) against your Genkit instance. Lighter alternative to the MCP route.

CrewAI tool for the Ejentum Reasoning Harness. Library of 679 cognitive operations across 4 harnesses, each engineered in two layers: natural-language procedure + executable reasoning topology (graph DAG). Injected before the LLM step to harden reasoning against decay on complex tasks.

LangGraph.js / LangChain.js integration for the Ejentum Reasoning Harness. createEjentumTools() returns four agent-callable tools as an array for createReactAgent / createAgent. Lighter alternative to the MCP route.

AutoGen integration for the Ejentum Reasoning Harness. ejentum_tools() returns four async tool closures (reasoning, code, anti-deception, memory) bound to shared config. Pass to AssistantAgent(tools=[...]).

smolagents Tool subclasses for the Ejentum Reasoning Harness. Four tools (reasoning, code, anti-deception, memory) plus an ejentum_tools() factory. Each call returns a task-matched cognitive operation engineered in two layers: natural-language procedure plus executable reasoning topology.

n8n community node for the Ejentum Reasoning Harness. Four operations (reasoning, code, anti-deception, memory) over 679 cognitive operations engineered in natural language. Marked usableAsTool for n8n AI Agent.