MCP servers that give coding agents real developer tooling: linters, package registries, test runners, language docs and local build commands. 13,967 servers in this category.
Provides recipes and working examples for running AI agents within isolated, secure NeevCloud Linux sandboxes.
Executes Python code via stdin to prevent shell escaping issues and allows direct access to a project's local virtual environment.
A self-hosted platform that manages multiple MCP backends as plugins with individual OAuth-protected endpoints and a web-based configuration UI.
Enables AI agents to inspect, replay, and mock mobile app traffic by connecting Charles Proxy to MCP clients.
Enables coding agents to perform agentic software engineering on Windows through surgical file editing, code search, and background job execution.
Provides a long-term knowledge system and retrieval index for AI coding assistants using Markdown notes and a Neo4j graph.
Runs a fixed suite of questions against an MCP server to score model answers and enforce deterministic tool-use assertions.
Analyzes MCP servers and agent skills to detect discrepancies between claimed capabilities and actual implementation risks.
Audits MCP servers to detect tool poisoning, hidden instructions, dangerous capabilities, and unconstrained inputs before AI agents connect.
Provides hardened terminal execution and file organization capabilities through a shared, deny-by-default safety core.
Lints and benchmarks MCP tool definitions to prevent LLM agents from selecting the wrong tools or constructing invalid arguments.
Decomposes natural language feature requests into task DAGs and executes them concurrently across isolated git worktrees with automated quality gates.
Provides AI agents with scoped git and GitHub operations including branch management, pull requests, and CI checks with built-in security guardrails.
Statically scans MCP servers and Claude Code projects to detect tool-poisoning, command injection, risky permissions, and leaked secrets.
Provides a private, self-hosted memory implementation that allows AI agents to search, read, write, and audit personal data via the Model Context Protocol.
Provides local-first conversation memory and evidence retrieval for AI coding assistants using a searchable corpus and RAG.
Provides AI assistants with a guarded local workspace to inspect, profile, and query analytical datasets like CSV, Parquet, and SQLite.
Expose OGC WMS, WMTS, and WFS geospatial services as tools for LLM agents to discover layers, render maps, and query vector features.
Provides a local hybrid retrieval layer for AI coding hosts to manage project memory that exceeds the model's context window.
Converts user-reported bugs into privacy-safe Playwright regression tests and structural debugging evidence.