dreamlake-ai

Create or repair HTML, React, Markdown, SVG, Mermaid, or code files for DreamLake’s isolated artifact frame.

Inspect a reachable dataset, author its .dreamrc, bind fields to views, and validate the result.

Prepare data for a DreamLake source, connect external storage, upload to an existing managed source, and verify the stored paths.

Use a local Python simulation rollout to write an indexed MCAP with Foxglove transforms, robot geometry, and metrics.

Prepare the fixed video subtask-labeling workflow from connected-source video and reference annotation paths.

Write or read annotations with the DreamLake Python SDK, using the video preset or a custom schema.

Publish a renderable file with the DreamLake CLI, add versions, list artifacts, manage sharing, or restore and remove one.

Generate a WorkflowSpec v1 JSON file from a data-production, labeling, or training goal.

Validate and publish a prepared WorkflowSpec v1 JSON file with the DreamLake CLI, then report its version and canvas link.

Verify that a connected DreamLake source contains each file a remote workflow will read.