Collection · 16 entries
MCP servers and Model Context Protocol tools
Production MCP servers, gateways, frameworks, and clients - everything in this directory that speaks the Model Context Protocol.
MCP went from spec to ecosystem fast. The tools here split into a few clear shapes: servers that expose data or actions to agents (Prometheus, Figma, codebase intelligence), gateways and aggregators that consolidate many servers behind one endpoint (1mcp), clients and CLIs that let you call MCP tools from a shell (mcpc, mcp2cli), frameworks for building production servers (Golf), and benchmarks for stress-testing them (MCPMark).
If you're new to MCP and want a starting point, look at the codebase intelligence ones (Qartez, codebase-memory-mcp) - they're the most concrete demonstration of what an MCP server actually changes about the agent loop.
Qartez MCP - codebase intelligence for AI agents
Rust MCP server that pre-indexes a codebase with PageRank, blast-radius, git co-change, and complexity signals. Agents query the graph instead of grepping files.
AgentShield - security scanner for AI agents
CLI, GitHub Action, and GitHub App that scan agent configs, MCP servers, and tool permissions for vulnerabilities. Detects skill poisoning and prompt-injection vectors.
Open Cowork - desktop app for AI coding agents
Open-source Windows/macOS desktop that one-click installs Claude Code, MCP tools, and Skills with sandbox isolation, multi-model support, and Slack/Feishu hooks.
AIMock - mock everything your AI app talks to
One package, one port, zero deps to fake LLM APIs, MCP servers, A2A, AG-UI, vector DBs, and search backends in tests.
Oxideterm - all-in-one terminal workspace
Native Rust + Tauri 2 terminal with local shells, SSH, SFTP, remote IDE, AI agent, MCP, RAG, and a file manager - all in one binary. No OpenSSL dependency.
1mcp - unified MCP server aggregator
Aggregates many MCP servers behind one endpoint. Acts as an MCP gateway/proxy so clients only configure a single server.
Prometheus MCP server
MCP server that lets agents query and analyze Prometheus metrics through a standardized tool interface. Useful for ops-aware coding agents.
mcpc - universal CLI for MCP
MCP client with persistent sessions, stdio + HTTP transports, OAuth 2.1, JSON output for code mode, and a sandbox proxy. Calls any MCP server from a shell.
MCPMark - stress-testing MCP benchmark
Benchmark harness that evaluates models and agents on real-world MCP usage. Comparable scores across servers and frontier models.
mcp2cli - turn any MCP/OpenAPI/GraphQL server into a CLI
Runtime adapter that exposes any MCP, OpenAPI, or GraphQL server as a flat CLI. Zero codegen, zero rebuild - handy for shell scripts and agent toolchains.
awesome-harness-engineering
Curated list for AI agent harness engineering: tools, patterns, evals, memory, MCP, permissions, observability, and orchestration.
AgentKit - multi-agent networks in TypeScript
Inngest's framework for multi-agent networks with deterministic routing and MCP-native tooling. Picks up where the LangGraph/CrewAI camps trail off.
Amazon Q Developer CLI
Agentic chat for the terminal - natural-language workflows wired into shell, MCP servers, and AWS tooling. Linux/macOS native, Rust binary.
codebase-memory-mcp - fast code intelligence MCP
MCP server that indexes a codebase into a persistent knowledge graph in milliseconds. 66 languages, sub-ms queries, single static binary.
Golf - production MCP server framework
Build, deploy, and scale MCP servers with auth, observability, debugging, telemetry, and a runtime baked in. Skips the per-server boilerplate.
figma-mcp-go - Figma MCP for free users
Go-based MCP server that gives agents read/write Figma access without rate limits. Text-to-design and design-to-code in one binary.
Frequently asked
What is the Model Context Protocol?
MCP is an open protocol from Anthropic that standardizes how LLM applications connect to external data sources and tools. An MCP server exposes capabilities (read data, run commands) that any MCP-aware client can call without per-integration glue code.
Do I need to build my own MCP server?
Often no - start with existing servers (codebase intelligence, Prometheus, Figma) and aggregators (1mcp). Build your own when you have proprietary data or a domain-specific action that no off-the-shelf server covers.
How do I evaluate which MCP server actually helps my agent?
MCPMark benchmarks server-and-model pairs on real-world tasks. For your own setup, the practical signal is whether the server reduces tool-call rounds or context bloat versus a baseline of grep/curl.
Related collections
Claude Code tools, plugins, and integrations
The best tools, MCP servers, and harnesses for getting more out of Claude Code - orchestration, observability, telemetry, and remote control.
Security tools for AI coding agents
Sandboxes, scanners, proxies, and governance toolkits that keep autonomous agents from doing damage.