MCP servers that give coding agents real developer tooling: linters, package registries, test runners, language docs and local build commands. 13,961 servers in this category.
Indexes Python codebases into a local SQLite database to provide agents with compact context and reduced token usage.
Connects ChatGPT Web to a local Windows folder to read Markdown files and images and submit generated reports back to a local directory.
Provides a secure framework for hosting and deploying custom Claude connectors as isolated, containerized tools with built-in guardrails and human approvals.
Provides local context infrastructure for AI agents to track folders and expose them as path-oriented, agent-readable content.
Lets an AI client classify prompts and route Manus AI tasks to cheaper models or chat mode to cut credit usage.
$claude mcp add credit-optimizer-v5 -- uvx mcp-credit-optimizer
Lets AI agents read and control desktop app interfaces through native accessibility APIs and Chrome DevTools Protocol.
Provides hands-on learning materials for local LLMs, RAG, AI agents, and MCP to help users build and verify AI-driven applications.
Converts any MCP server into a native CLI to reduce AI agent token consumption and enable Unix-style tool composition.
Provides a collection of community-contributed AI coding skills for automated software development tasks across various categories.
Enables coding agents to build, repair, migrate, and verify Model Context Protocol implementations through specialized engineering skills.
Automate Open 3D Engine with AI through editor control, project management, and asset pipeline tools.
Enables AI clients to develop, manage, and maintain Yandex DataLens dashboards through typed operations like reading, authoring, and publishing.
A hands-on workshop for building a multi-agent travel assistant using Microsoft Foundry, featuring MCP integration, RAG, and durable workflows.
Exposes MiniZinc constraint solving and optimization to LLM clients to parse, type-check, and solve models directly from a chat session.
Build Huawei eNSP network topologies and configure lab devices via console ports using an AI client.
Converts OpenAPI schemas into production-grade MCP servers with automatic tool generation, credential isolation, and built-in resilience.
Enables AI clients to execute code on Google Colab GPU runtimes directly from an Android device via Termux.
Enforces a structured DAIC workflow for Claude Code to prevent unaligned AI implementation through runtime hooks and protocol enforcement.
Provides a local, LLM-agnostic agent infrastructure with memory layers and persona-based system prompts.
Provides an active memory layer for AI agents to debate, cross-reference, and evolve structured project requirements and specifications.