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.
-
llm-pricesgithub.com/benbencodes/llm-pricesupdated this year
Zero-dependency Python CLI + MCP server for comparing LLM API costs across 144 models and 22 providers (OpenAI, Anthropic, Google, Mistral, xAI, DeepSeek, Groq...)
MIT ★ 2 -
Title and abstract screening for systematic reviews: zero-shot LLM classifier ranking with a statistical stopping criterion. Web version, Python package and MCP server.
Apache-2.0 -
The Karpathy LLM Wiki, production-ready. Zero-RAG personal knowledge base with MCP server, multi-agent support, knowledge graph visualization, and hallucination enforcement. No embeddings, no vector DBs — just markdown + git.
★ 5 -
zero-ai-gatewaygithub.com/vishtechie07/zero-ai-gatewayupdated this year
Zero-Trust AI Gateway: Dual-pass PII redaction (Regex + LLM), real-time security monitoring with Kafka/WebSockets, and MCP-powered compliance auditing for enterprise AI safety.
MIT -
Autonomous Zero-Bloat Agentic AI OS | 10 Core MCP Servers • 67 Production Tools • 14 LLMs • Cryptographic Provenance Receipts (HMAC-SHA256) • 3-Tier Self-Healing Memory • Sisyphus Orchestrator • 100% Passing
MIT ★ 2 -
Zero-trust AI control plane: strict stdio MCP server, MCP security gateway (JWT + RBAC), streaming PII guardrail, and a rate-limited failover LLM router, with an end-to-end demo and 348 tests.
-
Private, offline chat client for uncensored local LLMs. Zero dependencies, 228 tests, MCP server. Node + Ollama.
MIT -
@mempalace/coregithub.com/adshaa/mempalacejsupdated this year
Native TypeScript AI memory system and MCP server (Zero-LLM, LanceDB)
npm ISC -
@iflow-mcp/adshaa-mempalacejsgithub.com/iflow-mcp/adshaa-mempalacejsupdated this year
Native TypeScript AI memory system and MCP server (Zero-LLM, LanceDB)
npm ISC -
Token-budgeted codebase context packer for LLMs: XML, Markdown, llms.txt formats. CLI, MCP server, and GitHub Action. Zero runtime dependencies.
Apache-2.0 -
LLM agent that runs zero-knowledge proofs in plain English, with skeptical MCP tools that check, fill in or override the model's tool arguments. +23 points task completion across 3 model families. Code, data, paper.
MIT ★ 1 -
LLM-free, deterministic, explainable enterprise data retrieval over the Model Context Protocol (MCP). Zero tokens, zero embeddings.
-
fine-tuning-osgithub.com/Casius999/fine-tuning-osupdated this year
Zero-Data fine-tuning operations MCP server - 64 tools for the full LLM fine-tuning delivery lifecycle, callable by Claude Code
Apache-2.0 -
A from-scratch Model Context Protocol server (stdio, stdlib-only, zero deps): safe calc (OWASP LLM06), BM25 search, a no-LLM-judge answer grader, and live model-drift / eval-run lookups. 42 tests.
MIT ★ 1 -
Jev-powered search infrastructure for AI agents: zero API keys, MCP-ready, LLM-context aware, with local neural evidence verification.
pypi MIT ★ 20 -
kv-agent-visiongithub.com/Kvxw1105/kv-agent-visionupdated this year
Give pure-text LLM agents eyes: zero-dependency image understanding, OCR & image Q&A via MCP server + standalone Python script. 给纯文本 AI Agent 的图片理解能力(看图/OCR/图片问答)。
MIT -
mythos-mcpgithub.com/sorryorc/mythos-mcpupdated this year
Deterministic verification gate for Codex / Claude Code. 19 tools, zero LLM and zero network by default.
MIT ★ 1 -
tokenshrinkgithub.com/lhr-present/tokenshrinkupdated this year
Compress LLM prompts 30-60% — CLI, Claude Desktop MCP, browser extension. Zero API calls. Zero cost.
★ 2 -
goldbean-mcpgithub.com/wuzenghai616-lang/goldbeanupdated this year
GoldBean MCP Server v2 — Stateless, Edge-Ready. 48 Baidu AI tools (OCR/NLP/Vision/Speech/Translation/Face/Body/Creation/Search/LLM) via HTTP MCP. Zero handshake, zero session, zero SDK dependency. Compatible with MCP 2026-07-28. Deploy to Cloudflare Worke
npm MIT -
@agenson-horrowitz/agent-output-guard-mcpgithub.com/agenson-tools/agent-output-guard-mcpupdated this year
Agent Output Guard MCP Server - Validate and verify data from other agents before acting on it. Zero LLM costs, pure computation. Solves coordination failures in multi-agent systems.
npm MIT -
Fully local, privacy-first autonomous AI agent for laptop management — LangGraph reasoning loop, Ollama local LLM, ChromaDB RAG, MCP server, SHA-256 chained audit log. Zero API keys. Zero cloud calls.
Apache-2.0 ★ 5 -
Benchmark AI model inference speed — like speedtest.net but for LLMs. MCP server with zero telemetry.
MIT -
MCP server that routes LLM calls through CLI subscriptions (Claude, Gemini, Codex, Copilot) — zero API tokens needed
MIT ★ 2 -
A self-learning web scraper that uses LLMs to generate CSS extraction schemas once per page template, then caches and auto-generalizes URL patterns. Zero manual selectors. Zero maintenance. FastAPI + MCP server included.
MIT -
The high-performance, local-first data engine for developers and AI agents. Format, validate, and optimize data for LLM contexts without ever leaving your browser. Zero-server. Zero-leak.
MIT ★ 4 -
MCP server + Claude skill to control Spotify and build LLM-reasoned playlists via the spotify_player CLI (zero OAuth).
MIT -
llm-cachegithub.com/solarluminea/llm-cacheupdated this year
Drop-in LLM response cache — zero deps, streaming, fuzzy+semantic matching, Redis, MCP server, Express middleware
-
aiscrubbergithub.com/prvthmpcypher/aiscrubberupdated this year
Zero-telemetry browser-local privacy desk, CLI, and MCP server to scrub secrets before sharing with LLMs.
npm MIT -
Zero-dependency MCP server and CLI that puts Codex, Claude Code, Grok, Cursor, Antigravity, Pi, OpenAI and local LLMs behind one interface
Apache-2.0 -
Auto-discover llms.txt from your deps. Docs your agent can read. Zero setup.
pypi Apache-2.0 ★ 1 -
Brain-inspired long-term memory for AI agents — zero LLM during ingest or retrieval
pypi AGPL-3.0-or-later -
ComparEdge LLM Costno update data
Token cost math for LLM API calls: current per-million-token rates for 69 models across 17 providers, verified and date-stamped. Estimates single calls, comparisons across models, and monthly budgets. Zero dependencies, no API key.
smithery -
Know which free LLM APIs still work today — live free-model radar, change diffs, fallback chains, and an MCP server for coding agents. Zero-dependency Python.
MIT -
context-skeletongithub.com/msgtorahul-art/context-skeletonupdated this year
Zero-dependency structural code folding CLI, MCP server & LLM token optimizer
-
Keep local-model chat sessions alive indefinitely. MCP server giving any local LLM automatic context compaction and permanent memory. Zero dependencies.
MIT ★ 1 -
ragsync-mcpgithub.com/jsbroks/ragsync-mcpupdated this year
Give LLMs semantic search over your own docs — zero code, just a YAML config.
pypi -
PHP-MCP-Web-Toolsgithub.com/dmeldrum6/PHP-MCP-Web-Toolsupdated this year
A Model Context Protocol server for local LLMs — runs on shared hosting, zero dependencies.
MIT -
LLM-backed API mocks: every protocol, stateful, zero-edit HTTPS proxy, agent-native (MCP)
cargo Apache-2.0 -
A firewall for LLM traffic: detect prompt injection, jailbreaks, and secret/PII leaks. Zero-dependency Python library + CLI + MCP server.
MIT ★ 1 -
mcprunegithub.com/hamr0/mcpruneupdated this year
MCP middleware that prunes Playwright accessibility snapshots for LLM agents. Zero ML, 75-95% token reduction, all refs preserved.
npm MIT -
Zero-LLM structural code retrieval for AI coding agents, served over MCP
pypi MIT -
Deterministic zero-I/O utilities for LLM agents over MCP. A single dependency-free Rust binary.
MIT -
The secret manager for LLM agents: create, rotate, and use secrets they can never read. OS-native keyvault (macOS Keychain) + zero-dependency MCP server + Claude Code guard hook. No get verb, by design.
MIT -
Test impact analysis and code intelligence for LLM agents. Zero dependencies. 15 MCP tools.
MIT ★ 8 -
knowledge-curator-mcpgithub.com/Ryu07-d/knowledge-curator-mcpupdated this year
Local, zero-cost MCP server that fact-checks Markdown/Obsidian notes using a local LLM (Ollama) + free sources (Wikipedia, DuckDuckGo)
MIT -
agent-wiki-kitgithub.com/AvalancheAI-labs/agent-wiki-kitupdated this year
wikikit — one zero-dependency engine for every LLM wiki: lint (contracts, links, freshness), publish (llms.txt/HTML), serve over MCP, agent-driven ingest
MIT -
Zero-dependency LLM red-team scanner: 24 attack probes across 9 classes (44/20 in PRO), honest refusal-aware verdicts, Markdown report, x402 USDC + Nano paid tiers
MIT -
Local Laya System-1 judge MCP server + skill for oh-my-pi: typed decisions in milliseconds, zero LLM tokens
NOASSERTION ★ 13 -
sincegithub.com/LNSHRIVAS/sinceupdated this year
Stale-file detection and temporal context for LLM agents. Zero dependencies, any MCP client."
pypi MIT ★ 1 -
Connect RStudio to Claude Code, Codex, Gemini, and other LLM agents via MCP. Multi-agent orchestration, automated manuscript auditing, and zero-config setup with uvx
NOASSERTION -
Reversible, structure-aware context compression engine for LLM agents (Rust) — cuts token cost of tool outputs, JSON and logs. Deterministic, sans-io, zero network.
npm Apache-2.0 -
Graph-structured tool retrieval for LLM agents — zero-dependency, ontology-aware hybrid search
pypi MIT -
pbi-contextgithub.com/osc2405/pbi-contextupdated this year
AI context engine for Power BI models: indexed, queryable context for LLM agents (MCP server, CLI, docs), zero dependencies
pypi MIT -
@agenson-horrowitz/token-economist-mcpgithub.com/agenson-tools/token-economist-mcpupdated this year
Pre-execution LLM cost estimation, model comparison, prompt compression, and budget checking for AI agents. Zero LLM costs.
npm MIT -
Zero-dependency MCP server and SDK for deterministic PII and secret scrubbing with reversible LLM tokenization
npm MIT -
MCP server: ask a public GitHub repo questions, get line-cited answers. TF-IDF retrieval, zero-cost by default, optional LLM synthesis.
MIT -
Open-source, local-first long-term memory for AI coding agents: one MCP tool (recall) over a private local SQLite vault — zero cloud, zero vectors, zero background LLM calls, ~30 ms lookups. Apache-2.0.
Apache-2.0 ★ 1 -
Standalone belief and memory engine for Node.js — SQLite, zero dependencies, with a built-in LLM tool facade
npm MIT -
fetchmcpgithub.com/labtoolsstudio/fetchmcpupdated this year
Drop-in replacement for the official fetch MCP: clean, LLM-ready Markdown from any URL. JS rendering, anti-bot, honest errors, SSRF-safe. Zero-config, $0.
npm MIT -
🧠 Local-first LLM state branching & debugging. Capture, visualize, branch, and regression-test LLM calls as a DAG. Free & local via Ollama, any OpenAI-compatible API, MCP for agents. Zero config.
npm MIT ★ 3