SINT Labs builds open-source security infrastructure for AI agents that operate tools, data, robots, and industrial systems. Our public work focuses on runtime governance: capability tokens, policy gateways, MCP security, robotics safety, industrial automation controls, and tamper-evident execution evidence.
Website | Discord | X | Telegram
- AI agent security: pre-action authorization, typed deny/escalate outcomes, revocation, and audit evidence for autonomous workflows.
- MCP governance: security-first MCP proxying, tool-call interception, and evidence-backed policy decisions before tool execution.
- Physical AI safety: runtime controls for robotics, drones, smart homes, industrial automation, and other systems where actions affect the real world.
- Open standards evidence: conformance fixtures, public specs, compliance mappings, and submission packets for OWASP, MITRE ATLAS, AAIF, NIST, and EU AI Act-aligned governance.
sint-protocol- runtime security and governance for AI agents, MCP tools, robotics, industrial automation, and physical AI.sintos- core operating system and infrastructure for SINT AI agents.sint-discovery-cloud- trust and discovery layer for open-source and agentic work.sint-mcp-marketplace- stateless Hono MCP server deployable to Cloudflare Workers.twitter_mcp_server- MCP server for Twitter/X integrationsimage_mcp_server- MCP server for AI-powered image generation and editingdocs- SINT documentation and reference material
- Open-source tools for MCP security, agent infrastructure, and governed execution workflows.
- Protocol work that turns agent actions into scoped, auditable, and revocable operations.
- Practical examples and reference stacks for teams building beyond demos and into real-world deployment.
- Follow @sint-ai for new public releases
- Star the repositories you want us to invest in
- Open issues and pull requests when you find a gap worth fixing
Read CONTRIBUTING.md before opening issues or PRs.