MCP servers that give coding agents real developer tooling: linters, package registries, test runners, language docs and local build commands. 13,969 servers in this category.
Scaffold modern npm packages, CLIs, services, and applications directly through AI agents.
$claude mcp add packkit-mcp -- npx -y packkit-mcp
Enables AI assistants to debug Node.js applications using the Chrome DevTools Protocol with breakpoints, stepping, and variable inspection.
$claude mcp add devtools-debugger-mcp -- npx -y devtools-debugger-mcp
Enables AI agents to read files, edit code, run commands, and inspect git repositories directly on your local machine via a secure tunnel.
Records and interacts with browser pages, then exports polished GIF or MP4 demos with automatic zoom effects.
$claude mcp add pagecast -- npx -y @mcpware/pagecast
Enables coding agents to perform live X and web searches, conduct code reviews, and delegate complex tasks to the Grok Build engine.
Turns ChatGPT Web into a local development agent by providing repository access, task state, Git management, and command execution capabilities.
A modular agentic framework and MCP platform that lets you build autonomous agents and connect them to AI clients like Claude Desktop and Cursor.
Provides a local-first persistent memory layer that allows AI coding agents to share context, decisions, and preferences across different tools.
Provides Git-like undo and rollback checkpoints for OpenAI Codex CLI to snapshot, diff, and restore files before and after AI edits.
Delegates code reviews and bounded coding tasks from OpenAI Codex sessions to the local Grok Build CLI via an MCP server.
Coordinates multiple AI coding agents into a structured team that plans, builds, and reviews work using shared memory and a human-in-the-loop dashboard.
Allows AI coding agents to control a real Chrome browser to navigate, click, type, and debug web pages.
An engineering toolkit and MCP server suite that optimizes context usage, manages episodic memory, and enforces engineering discipline for AI coding agents.
Provides AI agents with a workflow for collecting primary sources, verifying provenance, and generating traceable research briefs.
Enables Codex to delegate parallel implementation, investigation, and testing tasks to persistent DeepSeek subagents with automatic completion callbacks.
$claude mcp add dsh-subagent-mcp -- npx -y dsh-subagent-mcp
Reduces model token usage through exact-recoverable context compression for terminal, file, project, web, and memory data.
Enables AI clients to manage project context and AI-readiness using the structured SKILL.md and project.faf formats.
Allows AI clients to control Termux on Android via localhost for command execution, terminal interaction, file management, and Android API integration.
Manages MCP servers within DeepSeek Harness by allowing users to monitor status, toggle servers, import configurations, and use AI to install new ones.
Enables AI clients to read Figma designs, extract design tokens, generate multi-framework code, and write changes back to the canvas via a companion plugin.