MCP servers that ship as code — a repository or a package you install and run locally (usually over stdio). They have no hosted endpoint, so we cannot probe or grade them: instead each one shows its repository, install command, license, stars and when it was last updated.
Collected from GitHub repositories and npm, PyPI, Cargo, Go, Maven, NuGet and RubyGems packages tagged as MCP servers, the official MCP Registry, Docker's MCP catalog, ModelScope, Baidu MCP World, Cline and other public directories; one entry per repository. A server that also offers a hosted endpoint is listed under Remote instead. These entries are not counted in our reachability, verification or readiness statistics.
-
mcp-xraygithub.com/tivaliy/mcp-xrayupdated this year
Xray MCP server: A Python service for proxying and managing Jira Xray requests with AI language models
MIT ★ 11 -
@aratiri/mcp-docs-proxystale
Proxy MCP server that exposes any llms.txt URLs as auditable documentation tools.
npm MIT -
Drop-in OpenAI / Anthropic / Gemini BaseURL Proxy & Edge MCP Server (~0.04ms). Cut LLM input tokens by 60–99% via AST code skeletons, JSON pruning & stream healing with 1 line of config.
MIT ★ 1 -
tare-aigithub.com/mstuart/tareupdated this year
Lossless-by-default context compression for LLM coding agents — HTTP proxy, CLI, and MCP server.
npm MIT -
HttpProxyMcpgithub.com/sailro/HttpProxyMcpupdated this year
An HTTP/HTTPS MITM proxy with full traffic capture, exposed as an MCP server so LLMs can inspect, filter, and analyze network traffic through natural language.
★ 14 -
taregithub.com/mstuart/tareactive
Lossless-by-default context compression for LLM coding agents — proxy, library, CLI, and MCP server. Local-first, cache-correct, reversible.
MIT ★ 5 -
markdowninthemiddlegithub.com/rickcrawford/markdowninthemiddleupdated this year
A local proxy that converts websites and APIs to clean Markdown. Convert HTML pages, JSON APIs, and dynamic sites. Get token counts for LLM budgeting.
MIT ★ 8 -
Keeps LLM prompt-cache prefixes stable so provider prompt caching actually hits: local proxy (point your client's baseURL at localhost) + MCP server. Fully local, no telemetry.
npm Apache-2.0 -
Request-path token proxy + analyzer for agent conversations. Live: cut 74% of input tokens on a real DeepSeek call (mcp-token-saver-proxy.fly.dev). MIT core, no-LLM, crypto-native Pro.
-
Cryptographic proxy and credential vending for LLM applications. Zero secrets on disk. mTLS everywhere. Apache 2.0.
Apache-2.0 ★ 1 -
Zero-dependency in-memory PII airgap & reversible masking proxy for Cloud LLMs (GDPR Art. 32 compliant MCP server)
Apache-2.0 ★ 4 -
sentinel-aigithub.com/MaxwellCalkin/sentinel-aiupdated this year
Real-time AI safety guardrails for LLM apps. 10 scanners: prompt injection, PII, harmful content, code vulnerabilities, obfuscation detection. Sub-ms latency. Python + TypeScript SDKs. MCP proxy. Claude Code hooks.
Apache-2.0 ★ 27 -
Self-hosted AI API key manager and LLM gateway. Store OpenAI, Gemini, Claude, OpenRouter and Groq keys once, call any model by name. Key rotation, OpenAI-compatible proxy, web dashboard, and an MCP server with sandboxed Code Mode for Claude Code, Cursor and Codex.
MIT -
ledgermindgithub.com/cognis-digital/ledgermindupdated this year
Local LLM cost & token forensics proxy with anomaly detection
NOASSERTION -
skill-forgegithub.com/gtm-k/skill-forgeupdated this year
Local-first skill runtime + visual manager for local/open LLMs — source agentskills.io skills and inject the right one at runtime into LM Studio, MCP hosts, and any OpenAI-compatible proxy.
Apache-2.0 -
fusion-mcp-hubgithub.com/KirtiJha/fusion-mcp-hubupdated this year
Solve the MCP token bloat problem. Build lean, custom MCP servers with only the tools you need from 300+ available tools. Reduce LLM costs and hallucinations. Includes CLI, WebSocket proxy, AI-powered recommendations, and interactive playground.
MIT ★ 2 -
token-enhancergithub.com/xelektron/token-enhancerupdated this year
A local proxy that strips web pages down to clean text before they enter your AI agent's context window. 704K tokens → 2.6K tokens. No LLM required.
MIT ★ 68 -
⚡ Autonomous AI Interception Proxy, Multi-Protocol OAST Canary Engine & Attack Hypothesis Ledger for LLM Pentesting Agents (Claude Code, Cursor, OpenCode, Hermes)
-
traffic_monitorgithub.com/pflaumax/traffic_monitorupdated this year
A FastAPI reverse proxy that streams traffic events through Kafka, exposes an LLM /ask endpoint that answers questions in plain language grounded in the live stats, and ships an MCP server.
-
token-scoutgithub.com/jackccrawford/token-scoutupdated this year
For OpenClaw, Hermes and more. Find free and low-cost inference (LLM models). Use them directly. Provides both a CLI and MCP server that knows which free-tier LLM APIs exist, which ones you have keys for, and which one fits your task. Returns endpoints so can you call models directly. No proxy, no middleware, no latency tax.
MIT ★ 9 -
Model Context Protocol (MCP) server & legacy HTTPS proxy bridging modern LLMs to OS X 10.7 Lion and dual-boot Linux.
MIT ★ 1 -
The local-first brain for coding agents: project memory + workflow intelligence + lossless ~50% context savings. MCP server + LLM proxy, zero cloud.
MIT ★ 6 -
An MCP server that provides tools for exploring large OpenAPI schemas
MIT ★ 14 -
Official MCP server for Proxidize — manage mobile & residential proxies from Claude, Cursor, and any MCP client. Rotate IPs, monitor usage.
npm MIT ★ 2 -
MCP server for NodeMaven residential proxies: geo-targeted proxy strings, sticky sessions, connection checks and usage
MIT ★ 24 -
pulsar-mcp-proxygithub.com/Keithw01/pulsar-mcp-proxyupdated this year
Public MCP server for Pulsar Developments — UK distributor of Mascot power supplies. Live GBP pricing, UK stock, competitor cross-references (Mean Well, RS PRO, XP Power, Phihong, CUI, TDK-Lambda, Victron +25 more), spec search, datasheets, EOL replacements, quote requests. Streamable HTTP: https://www.pulsardevelopments.com/mcp
MIT -
MCP server for AI agents: fetch web pages through residential proxies in any country. Works with Claude, Cursor and VS Code.
MIT -
Local MCP server that gives coding agents a verifiable project memory, trims what enters their context, and masks secrets — with honest accounting of what it saves and costs.
npm MIT -
Cut Claude Code costs: Claude stays the brain while its subagents run on cheaper models (OpenRouter, DeepSeek, Qwen, GLM, Kimi). Rust CLI, local proxy and MCP server with secret/PII redaction, budgets and savings reports.
MIT ★ 1 -
IcyCache - universal LLM gateway with a local cache. I see cache, I burn cache.
pypi MIT -
An MCP server with a built in lightweight proxy server & DB to capture & analyze network traffic within an AI agent
-
Short-term memory proxy gateway with proactive memory surfacing for AI agents
pypi Apache-2.0 ★ 5 -
MCP server + CLI for iOS Safari debugging via ios-webkit-debug-proxy — full WebKit Inspector Protocol support
MIT ★ 21 -
ATR: Automation for Token Reduction. An MCP proxy that hides tool lists and shrinks tool results so your AI uses fewer tokens.
pypi MIT -
Unified local gateway: LLM cache + compressed context drawer, in a single MCP server
pypi MIT -
Project-aware pooling proxy for Serena MCP servers: warm per-project instances with leases and an idle-TTL reaper (Windows-first)
MIT ★ 2 -
AgentCloakgithub.com/ryanfren/AgentCloakupdated this year
Secure email proxy for AI agents — content filtering, PII redaction, and prompt injection detection over MCP
NOASSERTION ★ 14 -
MCP server for Antigravity Claude Proxy — use Claude Opus 4.6, Sonnet 4.6, and Gemini 3 via any MCP client. TypeScript, tested, zero-config.
npm MIT -
agentsgategithub.com/agentsgate/agentsgateupdated this year
Undo for AI agents: an MCP proxy that checkpoints every tool call before it runs, so a bad `DROP TABLE` or `rm -rf` rolls back in seconds. Risk scoring, human-in-the-loop approval, tamper-evident audit log, dashboard, and MCP servers for PostgreSQL/MySQL/SQLite. Local-first, MIT, TypeScript, Node 20+.
MIT -
Smart Context Engineering Gateway for AI agents: lossless tool-output compression, episodic memory folding and a reversible diet_ref vault (MCP server + OpenAI/Anthropic proxy + local dashboard)
MIT -
llmtest-mcpgithub.com/tjacquesson/llmtest-mcpupdated this year
MCP server for LLMTest — benchmark AI models on real prompts, find cheaper alternatives across 340+ models. Works with Claude Code, Cursor, Windsurf.
MIT -
IntentShield for MCP: an offline-first security guardrail against prompt injection in MCP tool calls, with hooks for Claude Code, Codex CLI, Cursor and Gemini CLI and a transparent MCP proxy
pypi Apache-2.0 -
117 MCP tools for GPU infrastructure + Agent FinOps — deploy LLMs, manage VMs (Proxmox/XO/vSphere), track cost per agent, enforce budgets and model policies. Works with Claude, Cursor, n8n, LangChain
MIT ★ 20 -
ShadeNet: anonymous Tor egress for AI agents, zk-proof gated. A local HTTPS proxy and MCP server for the agent; Shade Tree nodes provide the egress. Research preview on Sepolia.
MIT ★ 9 -
Lightweight, single-binary, production-grade Rust LLM routing gateway & dashboard with health-aware failover, Claude Code compatibility, and MCP server.
Apache-2.0 ★ 4 -
Self-hosted AI workspace that compresses prompts before they leave your machine — 79% fewer tokens, measured. Drop-in gateway for Claude Code, Cursor, Aider and any OpenAI/Anthropic client, plus an MCP server. Free, MIT, zero telemetry.
MIT ★ 2 -
Coding-agent orchestration and MCP context runtime with governed repository execution, bounded tool results, and multi-agent session management.
npm MIT -
Russian Web Data Access for AI agents — remote MCP server (fetch_ru) + REST. x402/RUB.
-
One MCP server to aggregate them all - A unified Model Context Protocol server implementation
npm Apache-2.0 -
Share API keys without giving them away. Self-hosted key server and API proxy: give apps, AI agents, people and other servers their own tokens, with per-share keys, limits, timers, revoke and automatic key rotation. MCP, CLI and SDKs.
MIT -
SWAG reverse proxy configuration management via MCP. Create, edit, view, and manage nginx proxy configurations with auth integration.
MIT ★ 7 -
Shrinks your AI agent's tool output so it burns fewer tokens, without changing what the agent reads. Lossless by default, no decode step.
pypi MIT ★ 3 -
Hot-reload for MCP servers: edit, reload(), call again in the same session. Plus mount any of 130k+ community servers mid-session, Docker-caged.
npm MIT ★ 3 -
MCP-style HTTP server that multiplexes N database/REST backends behind one auth-protected endpoint - LRU cache, SQL allow-list, rate limiting, and per-target health checks.
MIT ★ 1 -
MCP-прокси для mcp.tutu.ru: каталог инструментов на 39% легче по токенам, детерминированная проверка ответов на выдумки, эвал-харнесс и трейс-вьювер. Хакатон Туту 2026
MIT -
Safety-first Database Guardrail & Sandboxing MCP Server for AI Coding Agents (Cursor, Claude Desktop, Antigravity). Protects databases from unconstrained mutations, drops, locks, and PII leaks.
npm MIT -
See what your MCP servers cost you in tokens — and cut it. Records every MCP tool call, itemises the per-session tax each server charges, and tells your coding agent what to remove. All local, zero config.
npm MIT -
AI Access Proxy Layer — granular permission gateway for AI agents. Connect services, define scopes, get an MCP endpoint.
MIT ★ 18 -
modelroutegithub.com/cognis-digital/modelrouteupdated this year
Local model router / proxy across Ollama, vLLM, and cloud with fallback
NOASSERTION -
Self-hosted MCP proxy for Basic Memory with explicit, fail-closed context routing and project lifecycle management.
MIT