Content-addressed code graph with 22 MCP tools for AI agents.
Tells you how you and your Claude agent actually work together. Local-only, no API keys.
Read-only pneumatic compatibility and evidence tools with stable IDs and canonical CompatAir URLs.
The Tenth Man Protocol — structured contrarian analysis for decision-makers via MCP.
Catch TypeScript API and OpenAPI contract drift in AI-generated code, then revalidate it in CI.
Catch bad animations before they ship. Deterministic motion audit + vision-LLM design review.
Read Project Gutenberg books via MCP. Search, fetch metadata, and retrieve text in passages.
TypeScript SDK for mutual ZKP authentication between humans and AI agents.
Native macOS MCP server for voice I/O — Swift binary, SFSpeechRecognizer + ElevenLabs TTS.
MCP server that converts raster images (PNG/JPG/WEBP) to SVG vector format.
Read-only MCP server over bounded.tools's Sigstore-verified static API (stdio).
Read-only MCP server over robertdelanghe.dev's Sigstore-verified static API (stdio).
Create a BoxTier-style PNG receipt from user-provided basketball game results.
Transcribe public videos & audio (YouTube, TikTok, IG) into accurate, timestamped text via API.
Supervised Meta Ads operating system for Claude Code with 57 campaign management tools.
Deterministic public-web change observation with evidence-bound commercial interpretation.
MCP server for controlling the Xojo IDE on macOS.
Deterministic Brotato theorycrafter MCP server: facts and math, not guesses.
Laravel documentation aggregator: multi-version core docs, services, packages, learning paths.
briefroom stdio MCP server — expose the briefroom CLI to Claude Code, Codex, and other agents.
Minimal streamable HTTP MCP server used for owned-account registry connectivity tests.
MCP server for x64dbg debugger - 23 mega-tools for AI-powered reverse engineering and debugging.
BrowserStack's Official MCP Server.
VirtualBox and VMware Fusion/Workstation behind one provider-neutral MCP tool surface.
About Development MCP servers
Development is the broadest category on mcp.site. It covers servers that give an AI agent hands-on access to the software lifecycle: sandboxed shells and terminals, version-specific library documentation, package registries, test runners, debuggers, build and deploy pipelines, and project tools such as issue trackers. Most run locally over stdio because they need your filesystem or toolchain; a growing share run as hosted endpoints you connect with an API key or OAuth.
When choosing one, start with what the agent should be able to do without asking you, then check how the server scopes that access. Good servers expose a small, typed set of tools, document every environment variable, and let you restrict paths or repositories. Prefer listings with an Official badge or a recently updated repository, and read the install snippet for your client before adding credentials.