Comprehensive risk detector with 12+ analyzers, 117 YARA rules, and ML-powered runtime monitoring to identify prompt injections, tool poisoning, rug pulls, and cross-server threats.

A Go-based analytics toolkit with OpenAPI-generated models, dependency injection, structured logging, and extensible tools, prompts, and resources for enterprise workflows.

ERC-1155 on-chain badges minted with verifiable context proofs and commitments, enabling multi-type badges with offchain metadata and secure reward capabilities.

Automates BitNet backend in Docker and exposes MCP tools to Copilot for VSCode.

AI-powered UI design selection with real-time comparisons, responsive previews, and HTML reference checks for complete, ready-to-use designs.

Nimiq MCP Server offers remote/local deployment with 18 tools, interactive elicitation, and edge-ready deployment.

Secure dynamic JavaScript execution with NPM package support and robust sandboxing, enabling AI-driven code experimentation with safe isolation and reproducible results.

A Java-based demonstration of the Model Context Protocol (MCP) using Spring AI. It showcases two transport servers that expose SivaLabs YouTube video metadata via STDIO and streamable HTTP, plus a client web app to query the data through MCP. Designed for quick experimentation, learning, and prototyping MCP integrations.

AutoMCP auto-generates Model Context Protocol tooling from OpenAPI specs, includes doc crawling, prompts, and multi-API deployment.

Playwright-powered MCP server that crawls Xiaohongshu posts to extract metadata and download videos and images, enabling targeted content insights, offline archives, and seamless MCP client integration. It leverages browser automation, supports share-link extraction, and returns structured metadata suitable for downstream processing and enrichment tasks.

Lightweight MCP server delivering detailed character breakdowns—total, spaces excluded, letters, numbers, and symbols.

Link AI clients to Fern's knowledge base for contextual assistance, docs, and SDK generation.

Two-package monorepo delivering native Flow tooling and DeFi/EVm-ready capabilities for seamless on-chain operations and cross-chain interactions.

Ruby-based Model Context Protocol implementation enabling secure, two-way data source integration with AI tools.

Python-based multi-agent orchestration for iTerm2 with persistent sessions, hierarchical teams, parallel layouts, and Claude Desktop integration.

PHP SDK implementing MCP to connect, expose prompts, and transport MCP messages securely.

Notion integration enabling AI models to query, modify, and organize workspace data through a standardized interface.

Spring Boot AI-based Confluence integration exposing document management as callable tools via @Tool, with ToolCallbackProvider registration and easy client testing.

Retrieval-driven refinement with data-derived intent and auditable prompts (single pass repair, contextual control).

Stateless bridge enabling Claude to fetch Swiss federal legislation from Fedlex via SPARQL metadata and HTML text, with multilingual support and no caching.

callmux enables batched, parallel, and cached tool calls with shared sessions and meta-tools to boost efficiency and resilience.

Dynamically exposes Nextcloud app APIs by OpenAPI discovery and proxies calls, adapting to installed integrations.

Claim-native memory with collision detection and hybrid search in a single-file SQLite store for AI agents.

Codebase intelligence tools for TypeScript projects, offering architecture overview, dependency maps, hotspots, and risk analysis to accelerate code understanding and safe refactoring.