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.
Provides a self-hosted gateway to connect MCP clients like ChatGPT or Claude to tools running on your local machines via outbound connections.
Turns codebases into queryable knowledge graphs using Neo4j and Tree-sitter to provide AI agents with exact call graphs and structural context.
Coordinates multiple AI coding-agent sessions on a single machine using a local, filesystem-based message bus.
Lets an AI client read and write Markdown files, evaluate computable document blocks, and render documents to PDF within a local workspace.
An autonomous QA agent that crawls web applications, infers test plans, and executes matrix test suites via MCP tools for AI coding agents.
An open-source browser agent that allows AI clients to control a real, logged-in Chrome browser through an MCP bridge.
Automatically discovers and loads MCP servers and slash commands from local configuration files into a dsh host with hot reloading.
Enables multiple AI coding-agent sessions to coordinate and exchange tasks via a local, event-driven message bus using the filesystem.
Enables AI agents to automate web browsers and perform API testing using Playwright and Chrome DevTools Protocol with visual element recognition.
$claude mcp add smart-browser -- npx -y @zhaolibin/smart-browser-mcp
Delegates text and code tasks to multiple AI models with automatic failover, shared mission memory, and specialized parallel subagents.
Allows AI agents to manage API lifecycles, including designing endpoints, running contract tests, managing mocks, and reading monitors via OpenAPI specs.
$claude mcp add routebase-mcp -- npx -y routebase-mcp
Connects Claude Desktop and other stdio-based MCP clients to a 1C database via an HTTP proxy.
Enables local agent-to-agent artifact exchange for LLM workflows through a shared catalog and versioned storage.
A vulnerable-by-design MCP server and Semgrep rule set for practicing and detecting object-level authorization bugs like BOLA and IDOR.
An automation agent that uses Playwright to perform human-like web interactions, network interception, and high-fidelity page observation.
Enables Claude to read, write, and manage Visual FoxPro projects, including DBF database inspection and code execution via COM automation.
Reduces MCP token costs by automatically generating optimized proxy modules that extract essential data from bloated API responses.
Orchestrates Claude Code and Codex CLI as parallel coding agents using git worktrees and a lead planner agent.
Manages MCP stacks, CLI tools, runtimes, and AI agent integrations through a unified local capability infrastructure.
Provides a coordinated engineering system for AI agents with specialized agents, skills, and memory for planning, testing, and reviewing code.