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

> ## Agent Instructions
> Start with /introduction/getting-started. Use the Inference API's Direct Query endpoint (POST /api/inference/query) with the newton-c-2.6 or newton-fusion-1.0 model (text, image, and video reasoning) or the omega-embedding encoder (time-series embeddings). ATAI_API_ENDPOINT is the base URL https://platform.u1.archetypeai.app with no version path; every endpoint path begins with /api. Pages whose descriptions are marked (Archived) document the legacy Lens runtime — do not use them for new projects.

# Label-Free Anomaly Discovery Agent

> Find rare anomalies in unlabeled sensor data, no normal reference needed.

> [!NOTE]
>
> This agent is currently in preview and is not available to all customers at this time.
> [Contact Sales](mailto:sales@archetypeai.team) for more information.

Finds anomalies in a single unlabeled sensor data file with no labels and no reference data of
normal operation. It treats the bulk of the file as normal and flags the windows that stand
apart from the rest. Every window comes back with a normal or anomaly label and an anomaly
score, with optional per-sensor scores that show which sensors drove the anomaly.

> [!NOTE]
>
> This agent currently only supports sessions, and needs the entire file before scoring. It
> works best when anomalies are rare, occurring less than approximately 10-15% of the time.

### Use Cases

* **Find anomalies when no clean baseline exists:** Compares each window with the rest of the
  same file, so teams get results on historical data without first isolating a stretch of normal
  operation.
* **Screen historical runs before deeper analysis:** Scores an entire recording in one pass and
  flags the windows that stand apart, so engineers review a short list instead of the full run.
* **Bootstrap a reference for ongoing monitoring:** Separates the probably-normal windows from
  the flagged ones, so teams can build a clean reference and move to Anomaly Detection for live
  monitoring.
* **Triage complex alerts and events:** Focuses investigation on the signals that matter most,
  cutting an alarm storm down to its likely origin. Ships behind a per-agent flag, so
  attribution can be turned on or off per deployment.


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