> ## Documentation Index
> Fetch the complete documentation index at: https://docs.vijil.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Review Discovered Resources

> Read what a scan returned.

A scan returns [Resources](/concepts/discovery/resource). This page covers where to read them and what the verdicts mean.

## Two Inventory Views

**The Findings tab** lists every Resource across every [Target](/concepts/discovery/target).
Each row has a **Classification** column showing `AI` or `Non-AI`, and a dash where the
Collector could not decide. The **Type**, **Source**, and **Kind** headers each carry a
filter.

**A Target's findings** open from the **Targets** tab. Expand a Target's row, and the
Console splits the results into three groups:

| Group | Meaning |
| - | - |
| **`AI`** | Identified as an AI or generative AI workload |
| **`Non-AI`** | Ruled out, with the reason recorded |
| **`Uncertain`** | The Collector could not decide |

Use the Findings tab for the whole estate. Use a Target's findings for one scope.

<Note>
  Discover counts findings per Collector. One model running on 200 laptops is 200 rows on the
  Findings tab, one per laptop, and 1 in the Resource total.
</Note>

## Open a Resource

Open a Resource from its identifier to see the evidence behind its verdict:

* Where the Collector saw it, and which Target and Collector kind reported it
* What answered, on which port
* What the Classification rests on
* Every scan that has observed it, with first and last seen

## Work Through Uncertain

A Resource lands in `Uncertain` for one of two reasons.

**The Collector had no language model configured.** The reason recorded on the Resource says
so. See [Classification](/concepts/discovery/classification).

**The evidence was thin.** This is common for network scanner Collectors, which see a port
and whatever answers on it. An endpoint or in-cluster Collector covering the same service
will usually settle it.

## Check What a Scan Could Not Reach

Every scan records its blind spots: subnets it could not route to, catalogues it could not
read, and hosts that gave no reply.

<Warning>
  An empty finding list has two possible meanings. The scope may be clean, or the Collector
  may have seen nothing. Read the blind spots first to tell the two apart.
</Warning>

Re-check blind spots after any access change. A Collector that loses a permission still
reports success, and its coverage shrinks.

## Keep Your Inventory Current

Set a recurring scan on the **Schedule** sub-tab of each Target you keep. **Last scan** on
the Targets tab shows how fresh each one is.

A registered Agent's card lists the Resources linked to it, with the time each was last seen.

## Next Steps

<CardGroup cols={2}>
  <Card title="Register a Discovered Agent" icon="bot" href="/owner-guide/discover/registering-agents">
    Move a finding into the Agent Registry.
  </Card>

  <Card title="Classification" icon="scan-search" href="/concepts/discovery/classification">
    How Discover reaches each verdict.
  </Card>
</CardGroup>


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.