AI & Agents MCP Servers for Codex CLI

MCP server for Movoice AI — create and manage voice AI agents directly from Claude or Cursor

MCP server for myCouncil — run multi-LLM debates from Claude Code, Claude Desktop, Cursor, and other MCP clients.

Zero-dependency single-file MCP server for Termux/Android, Raspberry Pi and Linux.

Verify a number before an agent asserts it — evals, backtests (Deflated Sharpe), signed receipts.

OmniCoreAgent is an open Python agent harness for building autonomous AI agents that use tools, manage memory and context, coordinate

Extended MCP server for OpenAI with GPT-5 model support, including codex variants

Complete MCP server for OpenAI Assistants API - Files, Vector Stores, Assistants, Threads, Messages, and Runs with 41 tools

9-LLM consensus + disagreement scoring + cheapest-route picks to fight hallucinations.

Production MCP server for multi-agent AI research with OpenRouter

MCP server exposing OpenRouter Fusion via async fusion_list/fusion_start/fusion_result with named configs (quality/budget/custom) —

AI model routing on your own vendor keys: pick the best model per prompt, or route and run it.

MCP server that makes PLAN.md the operational backbone of agentic development

Consistency verdicts for MCP tool calls: no two agents consume the same resource version twice.

Gamified Personal Prompt Coach — Duolingo for AI prompting with levels, streaks, and badges

Prompt quality analyzer — score, improve, and lint your AI prompts

AI-powered prompt refinement with adaptive questioning and multi-provider support (6+ LLM backends)

PromptLens — AI prompt analysis MCP server for Claude Desktop & Claude Code. Shows how AI reads your prompts with 5-axis scoring and

MCP server for Prova — let any AI agent verify its own reasoning before answering.

MCP server bridging LLM environments with Puter's free AI APIs

Qantara (arched bridge): an MCP bridge that lets coding agents (Claude Code, Codex, Gemini) delegate tasks to each other.

Find your AI agent's likely failure mode, get runtime settings, and clarify ambiguous prompts.