> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://dev.neurologyca.com/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://dev.neurologyca.com/_mcp/server.

# Overview

> Engine session workflow

The Engine API is used to process live video sessions and return insights generated by Neurologyca's internal models and proprietary insights pipeline.

The endpoints work together as a session-based workflow:

1. Call `POST /session` to create a new Engine session and receive the `sessionId` and Engine token for that session.
2. As the live video session is recorded, send `.kpn` packages to `POST /process-kpn`. Each package represents part of the session and provides the Engine with the data it needs to process that portion of the recording.
3. After the session has finished recording and all `.kpn` packages have been sent, call `GET /constellation/{sessionId}` to retrieve insights for the full session.

## Important notes

* Engine auth is separate from SoftSkills auth. Use only the token returned by `POST /session` for the Engine endpoints.
* For `POST /process-kpn`, enter the `sessionId` as digits only, without quotes.
* Fern does not automatically move values from one endpoint response into the next request, so you will need to copy `sessionId` and `token` manually.
* The exact internal structure of a `.kpn` package is proprietary and may evolve over time.