SODAX SDKDigital asset flows you can build and deploy with AI.
Promoted
Launching today
Aethra Runtime defence
Runtime defence and governance for AI agents
2 followers
Runtime defence and governance for AI agents
2 followers
Aethra AI is building a runtime defence and governance platform for AI agents - an enforcement layer that tracks data provenance and blocks untrusted instructions before they can drive agent actions. Our open-source SDK (Apache 2.0, Python/Node) is on GitHub, with an optional hosted control plane for approvals and audit. On AgentDojo, Aethra blocked 35/35 injection goals, 0 false positives (full LLM-backed evaluation in progress).
Hey Product Hunt
I'm Stanley, Co-Founder and Chief AI Researcher at Aethra AI, launching this with my co-founder Rishi Pal.
We built Aethra because AI agents are starting to take real actions — sending emails, executing code, moving data- and most teams have no way to stop an agent from being manipulated by something it reads, and no record of what it actually did. Aethra is a runtime enforcement layer that tracks where data comes from and blocks untrusted instructions before they can drive an agent's actions, in production.
We're early: working with design partners to pressure-test the architecture ahead of a public SDK release. We'd genuinely value feedback from anyone building or securing agentic systems, especially if you've hit this problem yourself.
Ask away below; happy to go deep on the architecture and what we're not solving yet.