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 a production-ready Python template for MCP servers featuring OAuth 2.1 authentication, tenant isolation, rate limiting, and audit logging.
Provides a multi-user gateway for Forgejo that enables hosted assistants like ChatGPT to interact with repositories via OAuth or token pass-through.
Visualizes the step-by-step execution of an AI agent loop, showing thoughts, MCP tool calls, and results on a live timeline.
Demonstrates native LLM tool calling through runnable examples including single-step, multi-step loops, inventory assistants, and MCP-hosted tools.
An AI agent that interacts with local documents, weather services, and inventory databases using Ollama or Claude via the Model Context Protocol.
Provides a collection of testable, security-focused MCP servers for knowledge base retrieval, sandboxed file management, and web fetching.
Enables semantic search over Obsidian vaults and delegates low-stakes tasks to local LLMs via Ollama to preserve context and privacy.
A Python-based MCP client that enables AI clients to list servers, call tools, and read resources via stdio or HTTP transports.
An open-source AI agent that uses a guarded GitHub MCP server to read issues and pull requests with restricted commenting capabilities.
Provides a normative specification and a linting tool to standardize tool naming, grouping, descriptions, and discovery for MCP gateways.
Provides local persistent memory for Claude Code and Claude Desktop using a SQLite database to store facts, preferences, and session states.
Build production RAG pipelines and MCP servers using a modular Python framework with hybrid search and FastAPI integration.
Provides local-first code intelligence through hybrid keyword, semantic, and call-graph search for repository analysis and impact assessment.
Provides a governance layer for MCP servers featuring role-based access control, JWT authentication, PII redaction, and audit logging.
A production-ready Python toolkit featuring a guarded MCP server, a framework-free agent loop, and a versioned prompt registry with A/B evaluation.
Generates and maintains API connectors from OpenAPI specs, documentation, or MCP servers to enable standardized execution and testing.
Lets an AI client model, inspect, and export CAD geometry using Ansys SpaceClaim or Discovery through PyAnsys Geometry.
Provides long-term memory for AI agents by storing typed JSON objects in S3-compatible storage with BM25 and local embedding search.
Acts as a policy-as-code security gateway that enforces least-privilege tool policies, redacts sensitive data, and scans for prompt injections in MCP servers.
Executes model-written Python code against MCP tools within a secure, locked-down Docker sandbox using a host-side validation bridge.
