Launching today

Agent Black Box
Investigate AI agent actions with causal graphs
2 followers
Investigate AI agent actions with causal graphs
2 followers
Trace AI agent and MCP tool actions through causal graphs, authority analysis, and evidence-backed incident timelines. Agent Black Box is available as a hosted, invite-only Closed Beta V1 using synthetic demo data. The current release is observation-only. Request early access






Free Options
Launch tags:Security
Launch Team

Render Sandboxes: a computer for every agentRun untrusted code safely. Idle sandboxes cost nothing
Promoted
Hey Product Hunt! 👋
I built Agent Black Box because understanding what an AI agent actually did across tools, permissions, and downstream actions can be difficult.
ABB helps reconstruct those actions using causal graphs, authority analysis, and evidence-backed incident investigations.
Our hosted Closed Beta V1 is now live with synthetic demo data. It’s observation-only and available to approved testers.
I’d love feedback from AI agent developers and security engineers.
What evidence would you need when investigating a suspicious AI agent action?
Request early access: https://abbxai.com