MMCP ServersDeveloperlow

HighlyLoadedEgo/ClefMCP

Name: HighlyLoadedEgo/ClefMCP
Author: Various

by Various

Local MCP server exposing the Cloudflare Clef-Flash decision model to AI agents — structured decisions with probability outputs, fully offline via llama.cpp.

Visit site Submit your build →

MCP

HighlyLoadedEgo/ClefMCP

Added 18 Oct 2026

#ai-agents#clef#decision-model#llama-cpp#local-llm#mcp#model-context-protocol#offline-inference

Catalogue record, not a reviewed recommendation. This record can include automatically collected descriptions and setup details. Confirm important claims with the original source. Explore source-reviewed tools.

Overview

Local MCP server exposing the Cloudflare Clef-Flash decision model to AI agents. It provides structured decisions with probability outputs and runs fully offline via llama.cpp. Written in TypeScript, it is a developer tool for integrating the model into agent workflows.

Best for

Best for
Developers who need a local, MCP-compatible decision model for AI agents

Use cases

Give an AI agent a structured decision-making endpoint with probability scores
Run the Clef-Flash model locally for offline agent reasoning
Wire Clef-Flash into MCP-compatible clients for decision support

Notes

7 stars on GitHub. Last updated 2026-10-04. Licensed Apache-2.0.

Pros

Fully offline execution removes external API dependencies
Probability outputs give agents calibrated confidence information
Simple MCP integration for existing agent tooling

Cons

Very early-stage project with only 7 stars
Depends on local llama.cpp setup and model management
Limited community adoption and documentation

Free 27-page guide

Get the free Developer’s Field Guide

A 27-page field guide to the AI coding workflow with Claude. Claude Code, MCP servers, the prompt patterns that work, and what to delegate. Free.

Enter your work email. We send it straight over, plus a few short notes worth knowing. Unsubscribe any time.

← Back to MCP Servers Submit your own entry →