AI & ML MCP Servers
LLM providers, agents, models, embeddings, inference, and machine learning workflows.
Morse — Telegraph in Telegram
Published by GitHub account Harshyadav442277
Hosted MCP endpoint that sends questions to Telegraph's ranked miners and returns answers with a verifiable receipt, no wallet needed.
RAGScore
Published by GitHub account HZYAI
RAGScore is a local MCP server that generates QA datasets from your documents and evaluates a RAG endpoint, scoring answers and diagnosing failures.
Erabi
Published by GitHub account HMAKT99
Lets an assistant check paid x402 agent APIs for signed uptime, latency and price attestations, and join a pre-release agent intent and reputation exchange.
WhitePact
Published by GitHub account Guruprasath-Annadurai
WhitePact is an AI governance MCP server offering trust scoring, guardrails, bias and hallucination-risk signals, and compliance mapping for agent actions.
Woody Mcp
Published by GitHub account Gunsmith-Academy
Woody MCP gives an assistant a firearms knowledge base, jurisdiction-specific legal lookups, and specialist routing for build, ballistics, and maintenance…
GetABrain
Published by GitHub account Guitarmaniac24
GetABrain lets an AI assistant submit questions to real human workers, wait for structured answers, and rate them, using a prepaid balance.
QECTOR Quantum Error Correction
Published by GitHub account GuillaumeLessard
Local quantum error correction server for Claude that decodes syndromes and verifies each correction against the parity relation H c = s (mod 2).
Gemini Cloud Assist Mcp
Published by GitHub account GoogleCloudPlatform
Connects an assistant to Google Cloud's Gemini Cloud Assist Cloud Operator agent to investigate incidents, review rollouts, review Terraform and optimize…
Toolfactory
Published by GitHub account GoatInAHat
A scaffolder that turns one operation module into agent skills, MCP servers, plugins, CLIs and packages, and keeps every generated surface in sync.
GlianaAI
Published by GitHub account Gliana-Labs
Lets an assistant browse and run pay-per-call generative AI models (image, video, music, speech, chat) plus utility tools, paying per call from a local crypto…
Gildaraio
Published by GitHub account Gildaraio
A local MCP server that gives AI agents operating contracts: structured instructions compiled into system prompts.
Vaultfire Trust Protocol
Published by GitHub account Ghostkey316
Local MCP server for on-chain trust verification and "Street Cred" scoring for AI agents, spanning ERC-8004 and multiple chains.
PersonalKnowHow
Published by GitHub account Georgi-Petkov
Remote MCP server exposing one person's public learning and work history as a queryable knowledge graph of projects, courses, certifications and positions.
Easel
Published by GitHub account Gemberkoekje
A headless painting engine exposed over MCP, letting an assistant lay real brush strokes, inspect the canvas, and export a painting.
Garl Protocol
Published by GitHub account Garl-Protocol
GARL Protocol MCP server gives an assistant tools to verify agent trust scores, issue and check capability tokens, and resolve signed action receipts.
Agent Trust
Published by GitHub account Garl-Protocol
Gives an assistant tools to submit signed action receipts, look up agent trust scores, and issue or verify capability tokens for AI agents.
Agent Services Mcp
Published by GitHub account Gareth1953
A local MCP server that exposes three remote services for certifying content provenance, scoring content quality, and auditing agent actions, with signed Ed2559
Promoshot
Published by GitHub account GarAlex
Headless PromoShot engine that lets an assistant author, validate, inspect and render .promo video projects to stills, GIFs, frame sequences or mp4.
Gandr Mcp
Published by GitHub account Gandr-AI
Gandr's official text-to-speech server that lets an MCP agent render speech in 23 languages with six voices and check token usage.
GabrielaHdzMicrosoft
Published by GitHub account GabrielaHdzMicrosoft
Lets an assistant store and retrieve visual memory and context as a knowledge graph.
RECOVER — Decide Safely After a Failed Machine Action
Published by GitHub account GSterlingPress
A remote decision service that turns a failed API, tool, or agent action into one of six safe next steps before you retry.
POSTFACT — Resolve Ambiguous Side Effects Before Retry
Published by GitHub account GSterlingPress
Remote MCP server that tells an assistant whether an ambiguous side-effecting action actually completed, so it can decide whether a retry is safe.
ONCE — Exactly-Once Reliability for AI Agent Actions
Published by GitHub account GSterlingPress
Remote MCP service that wraps consequential agent actions with idempotency keys, post-state verification and durable receipts to prevent duplicate side effects.
Humanizer
Published by GitHub account GPTHuman-ai
GPTHuman server that rewrites AI-generated text into more natural-sounding prose, with tone and rewrite-mode options and detector metadata.