Samma Suit
Provides governance and security for autonomous AI agents with eight enforced security layers, including gateway protection, permissions, skill vetting, cost controls, containment, identity, audit trail, and kill switch.
Sammā Suit is an open-source platform that provides governance and security for autonomous AI agents, enforcing eight security layers across all interactions. It solves the problem of deploying AI agents safely in production by ensuring every message is protected against threats like prompt injection and unauthorized access. The platform targets developers and IT teams in mid-market and enterprise companies who need to manage AI agents with strict security and compliance requirements. It is positioned as a self-hosted, open-source solution that supports multiple AI providers and integrates with existing agent frameworks like OpenClaw.
Key features
- 8 enforced security layers
- 5 channel adapters
- 30 governed skills
- Multi-provider support (Anthropic, OpenAI, Google)
- Per-agent budget enforcement
- Ed25519 cryptographic identity signing
- Full audit trail with enforcement trace
- Kill switch with state management
- Skill marketplace with AST scanning
- Real-time cost tracking
- BYOK (bring your own API keys)
- Local LLM support (Ollama)
- No social media activity within the last 30 days
- Marketplace
- Community
- Docs
- Software developers
- Engineering teams
- IT teams