"FIRST" matching MCP servers:

Long AI conversations fail in predictable ways. Context-First fixes all four: Failure Mode What Goes Wrong Context-First Solution Context Drift AI forgets earlier decisions and intent as the conversation grows context_loop + detect_drift continuously re-anchor every turn Silent Contradiction New inputs silently overrule established facts — the AI doesn't notice detect_conflicts compares every inp

First Voice AI MCP for AI Agents

A Model Context Protocol (MCP) server that implements AI-First Development framework principles, allowing LLMs to interact with context-first documentation tools and workflows for preserving knowledge and intent alongside code.

Local-first AI agent for approval-gated automation and verifiable LLM workflows.

World-first unified MCP gateway + Control Barrier Function (CBF-QP) safety engine for agentic robotics, enabling LLMs to control physical robots with mathematically proven safety.

Offline-first MCP proxy with policy engine and immutable audit trail for secure AI coding agents.

Open-source AI marketing agent toolkit with plan-first workflow, human checkpoints, and BYOK support.

🧱 easy fast local-first microVM runtime and library

Enables any MCP-compatible AI agent to ask humans questions in real time via a mobile-friendly PWA with push notifications, operating local-first without third-party services.

Enables secure, privacy-first SQL database interaction via natural language, using local LLM and PostgreSQL with dynamic tool-calling.

this is my first mcp

Exposes the four Ejentum cognitive harnesses (reasoning, code, anti-deception, memory) as MCP tools any agentic client can call. Drop-in scaffolding that catches LLM failure modes like sycophancy, hallucination, and reasoning shortcuts.

Post AI-generated images to Vynly, the AI-only social feed. Four tools for posting, reading, searching, and ephemeral sparks — auto-claims a demo token on first run so it works with no signup.

Hand a project over — to your next session, to Cursor/Codex/Gemini, to your successor — with the reasons attached. Local-first memory MCP for coding agents: decisions re-injected before the agent acts, drift detection across every project you run. One SQLite file, nothing leaves your machine.

Local-first external brain for Claude Code, Codex, and any MCP client. Stores decisions, entities, and session artifacts in one SQLite file and exposes MCP tools for recall, page, promote, review, graph-query, and source-status.

Official MCP server for Algenta, a local-first AI engine powered by Mojo CPU/GPU kernels, 525+ Mojo libraries and 6,000+ compiled functions. Governed data discovery, exact queries, decisions, simulations, persistent decision memory and agent runs with approvals and execution receipts. Runs on your infrastructure. Source: packages/mcp/algenta_mcp. Install: pip install algenta-mcp (stdio).

Local-first shared memory and coordination layer for AI coding agents, with repository evidence, reservations, handoffs, code graph context, and dashboard review backed by PostgreSQL/pgvector.

A local-first MCP memory layer for coding agents that automatically organizes shared memory writes and provides a governance console for inspection, correction, and rollback.

A local-first middleware proxy that proxies multiple MCP servers through a single entry with schema minification, semantic routing, security hardening, and state guard.

Technical Preview: local-first decision memory for AI agents. Agents can keep an answer while losing why it was chosen, what was rejected, and what failed; ShadowGraph stores that reasoning and lets callers review decisions against stored facts. Local stdio MCP; install from GitHub with Node.js 20+.