A

ai-agents

MCP server for RiskState — pre-trade risk governance API for BTC and ETH crypto trading agents. Spot, perps, and DeFi aware.

Truitr is an AI-powered social network where users and autonomous agents can create content, interact through posts, comments, reactions and direct messages, and communicate across multiple social network instances through interoperability APIs. Built with Django, PostgreSQL, Azure AI, WebSockets and Docker.

Virtual Company Kernel is a public framework for running software delivery with AI agents under clear governance, human approvals, reusable workflows, and auditable execution artifacts.

AI-powered security orchestration for GitLab CI/CD. Automated vulnerability patching, threat modeling, and compliance scoring with GPT-4 and Claude AI agents.

Persistent and containerized development environment for Gemini CLI (https://geminicli.com/). This project provides a secure, automated workspace featuring built-in skill management (context-manager for memory, gema-manager for agent orchestration) and full support for MCP (Model Context Protocol). Designed for seamless local development, robust state persistence, and modular agent integration.

Use Case Study: Create Agentic IA With Strans Agents SDK and AgentCore Runtime on AWS

MCP-native browser automation server — 62 composite tools, 153 CLI subcommands, zero config. Give AI agents full browser control via the Model Context Protocol. Navigate, interact, extract text/HTML/screenshots, debug network, manage sessions, visual diffing, React inspection, auth vault, HAR recording. Works with Claude, Copilot, VS Code, Cursor. Built on Playwright/Chromium. The most comprehensive browser MCP server available.

Coordinated Agent Swarm Testing — open-source framework that deploys 100 Claude AI agents in five specialized squads (Security, Logic, Robustness, Performance, Data Integrity) to systematically analyze codebases for bugs, security issues, and code quality problems. A research-backed substitute for human beta testing. Empirical results from the published paper: 100 agents analyzed a 12,500-line React/TypeScript application in 63 minutes, identified 687 unique findings, at a total cost of $7.97 — compared to $2,000-$20,000 for equivalent human beta testing. Available via npm (cast-swarm) and as an MCP server. Apache 2.0 licensed. Published preprint available.

Describe what you want. Go home. Sandcastle ships it. 6 AI providers, EU data residency, smart failover, cost intelligence, 20 step types, 236 templates. European-built, open source. pip install sandcastle-ai

Collection of purpose-built MCP servers for AI agent workflows.

Real-time debugging dashboard for AI-assisted development. Monitor live events, services health, code quality, and AI agent interactions through an intuitive web interface. Features SSE streaming, VS Code integration, and 13 interactive tabs for comprehensive project insights.