Find the AI Agents Behind Your Endpoints
You cannot secure what you cannot see. Discovery probes the HTTP endpoints you list, identifies the AI agents behind them and the provider that serves each one, and keeps them in one inventory you can re-check on demand.
In this release
You can use this in your workspace today.
- Asset mapSaved agents and the tools, datasets and vendors you declare, with their recorded relationships.Live
- ModelsThe provider and model recorded on each saved agent.Live
- DatasetsYour declared data sources and classifications, linked to saved agents.Live
- AIBOMExport the saved agents and declared assets visible in your current inventory view.Live
Endpoint Probing
Probe the HTTP endpoints you list and identify which ones are AI agents and which provider serves each one. Discovery does not sweep IP ranges.
Health Checks
Re-check every agent in your inventory with one on-demand health check, or check a single agent to see whether it responds and how long it takes.
Agent Inventory
Keep every agent a sweep finds in one inventory, with its provider, its model when the endpoint reports one, its endpoint and when it was last checked.
Risk Flags
Flag each discovered agent from what the probe sees: its provider, whether it requires authentication, whether it responds, its latency and whether its URL looks internal or external. A heuristic to prioritise review, not a verdict.
Shadow AI
Teams deploy AI models without central oversight. Point Discovery at the endpoints your teams run and it tells you which ones are AI agents, reducing blind spots.
Ungoverned Models
Agents deployed without security scanning create regulatory risk. Every agent Discovery finds lands in your inventory, ready to enter your governance pipeline.
Attack Surface Sprawl
Every unreviewed AI endpoint is an attack surface. Discovery maps the AI endpoints you point it at so you can prioritise protection.