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.
-
snapdrive-iosgithub.com/MasamiYamate/snapdrive-iosupdated this year
AI agents drive iOS Simulator to generate snapshot baselines and run visual regression tests.
MIT ★ 1 -
Autonomous Telegram-bot QA for Claude Code: snapshot-test dialogs over MTProto, specs from source code, zero tokens at run time.
MIT ★ 2 -
agentsnap-mcpgithub.com/MukundaKatta/agentsnap-mcpupdated this year
MCP server: snapshot tests for tool-call traces. Wraps @mukundakatta/agentsnap.
npm MIT -
blender-mcp-ocarinagithub.com/msradam/blender-mcp-ocarinaupdated this year
Automate and snapshot-test Blender through MCP with Ocarina. No LLM.
-
@lamalibre/install-portlama-e2e-mcpgithub.com/lamalibre/portlamaupdated this year
MCP server for Portlama E2E test infrastructure — VM lifecycle, snapshots, test execution
npm SEE LICENSE IN LICENSE.md -
Let AI agents see, drive and test your iced (Rust GUI) apps over MCP: snapshot, click, type, assert and record regression flows.
cargo MIT -
MCP server + agent skill that catches fake and fragile AI-written tests (no assertions, tautologies, swallowed failures, .only, sleeps, snapshot-only, mock-heavy), runs the real test runner and flags zero-test runs. JS/TS, Python, Go, Swift, Kotlin/Java.
npm MIT -
Drive a live Flutter app from an AI agent or CI: Semantics snapshots, gated taps and typing, screenshots, perf traces. No flutter_test harness, no test file between actions.
MIT ★ 3 -
React re-render profiler with snapshot testing — CLI, GitHub Action, MCP server and Agent Skill. Finds wasted renders and the exact line to fix.
npm MIT -
Drive an Android emulator as a deterministic test harness, from Claude Code. MCP server with snapshots, a compact screen index, and evidence-collecting assertions.
pypi MIT -
Local MCP server (stdio) that extracts live pages and multi-step flows into compact Semantic JSON with Playwright-native locators, observed behavior (navigations, requests) and snapshot diffs, for consistent AI-generated Playwright tests.
npm MIT -
allure-reader-mcpgithub.com/naelmarwan1/allure-reader-mcpupdated this year
MCP server that reads Allure test results — failure stacktraces, step details, screenshots and DOM snapshots — from a local directory or a Jenkins build artifact.
npm MIT -
dev.dungbeetle/mcpgithub.com/dungbeetletech/cloudno update data
Visual regression & snapshot testing for AI agents — list runs, read semantic diffs, review.
npm -
tui-tdgithub.com/vurte/tui-tdupdated this year
tui-td drives terminal applications in a PTY, captures ANSI state as structured data, and provides PNG screenshots and HTML renders. Includes an MCP server for AI-driven testing, a JSON test runner, RSpec matchers, Minitest assertions, semantic selectors, and named snapshot testing.
rubygems MIT -
TypeScript inferred types, completions and type errors for AI coding agents and type tests: an MCP server, a CLI, and snapshot-test helpers
npm MIT -
mcp-app-servergithub.com/sharathStack/mcp-app-serverupdated this year
FastMCP server integrating a stateful app into RL sandbox environments — populate/snapshot/restore hooks with content-addressed state digests, Docker packaging, and a 51-test suite spanning the app, state, and protocol layers.
-
Cinema 4D MCP — production fork (93 tools, +68 over upstream). Comprehensive Scene Nodes authoring/dissection (22 patterns, 802 templates, 40 verified labels), Octane OSL injection, fields, deformers, volume builders, scene snapshot/diff, recipe-based regression tests. C4D 2026 compatible.
MIT ★ 20 -
BannerlordInspectorgithub.com/davrodwconnections/BannerlordInspectorupdated this year
Read-only live inspector for a running Mount & Blade II: Bannerlord campaign - in-game module plus MCP server, 37 tools: swallowed-exception capture, Harmony conflict diagnosis, content auditing for total conversions, world snapshots and test checklists
MIT -
cdpilotno update data
cdpilot is a browser automation CLI and local MCP server that talks to Chrome, Chromium, Brave or Vivaldi directly over the Chrome DevTools Protocol, with no Puppeteer, Playwright or Selenium and no browser download. Its 45 MCP tools cover navigation, clicks and form filling (including inside cross-origin iframes), accessibility-tree snapshots, screenshots, test assertions, a page's own WebMCP tools, and Web Bot Auth request signing. `cdpilot connect` attaches to a browser you started yourself, so a person can clear a login or CAPTCHA and the agent continues; cdpilot never closes that browser. Every feature has a runnable example with its real output: https://github.com/mehmetnadir/cdpilot/tree/main/examples
smithery -
The StoryBook for native mobile applications
npm MIT ★ 14 -
Accessibility tooling that works in the real world — Extracts the accessibility tree as plain data. Use it through a Chrome extension, Storybook panel, React component, testing library, CLI, or MCP server — one engine, every surface.
npm MIT ★ 2 -
Playwright-style testing for Windows desktop apps — as a library and an MCP server, built for legacy (VB6/MDI) realities.
MIT ★ 3 -
Prove that what already worked still works after an agent changed the code. Picture checks, guards for fixed bugs, a pre-release walkthrough and known-good markers — as a CLI and as an MCP server.
npm MIT -
Rehearse SQLite migrations on a disposable snapshot. Verify rollback data and schema with a Java 21 CLI, offline reports and read-only MCP tools.
MIT -
MCP Toolkit for Flutter AI Agent Driven Development (MCP/CLI + custom client side tools) - via closed feedback loop (visual & semantic snapshot) and high client side customization adaptable for any Flutter app. Nowadays it is often called as agentic harness.
npm MIT ★ 398 -
Selenium MCP server for AI agents — 40+ browser automation tools with page snapshots, selector hints, multi-session, and batched execution.
npm MIT -
See and drive a GPUI app from the command line or MCP: snapshots with refs, clicks, typing, waits, screenshots and recordings
Apache-2.0 ★ 17 -
Branch SQLite like git — copy-on-write forks, checkpoint/rollback/promote, crash-tested capture. Stock SQLite files on local disk or S3. Built for AI agents and eval harnesses.
go Apache-2.0 ★ 6 -
Enhanced mobile app automation for AI coding agents. AI-powered selectors, optimized snapshots, streaming MCP, AI screen recording, security scanning, interactive architecture docs. Fork of callstack/agent-device.
MIT -
nextjs-agent-mcpgithub.com/zohaib3249/nextjs-agent-mcpupdated this year
MCP server to drive and inspect a running Next.js app: headless route/error introspection + a real-event in-page browser bridge (click, fill, snapshot, React fiber tree, network, storage, eval) — no Playwright.
npm MIT -
[PREVIEW — experimental, built for teaching & workshops; not for production.] A Model Context Protocol (MCP) server exposing a real browser (Playwright) as MCP tools — goto, click, type, snapshot, screenshot, evaluate, console/network logs, dialogs. Pure .NET, run via dnx — no Node, no npx.
nuget MIT -
MCP server that lets Claude Code, Cursor, Antigravity and other AI agents see and control your Expo / React Native app on a real phone - no more pasting screenshots. Snapshot the screen, tap, type, scroll, screenshot, read logs and network. Works with Expo Go, iOS and Android.
npm MIT ★ 2