Use with agents

Lightdrift is agent-native by design: requests are plain JSON, every answer is fully machine-readable, and the rights question, the thing that usually needs a human, is answered inline.

​

Connect your agent

Choose the connection for the job, or install both. Each guide follows the dashboard’s client setup options and adds verification and troubleshooting steps.

​

Patterns that work well

• Grounding generation — retrieve a real photo of an obscure subject first, then hand it to your image model as a reference. The generator stops guessing.
• Slide and page builders — describe the subject and the composition you need (“wide shot, open sky on the left”) and use orientation and min_width filters for the slot.
• Answering with real images — for anything past the canonical middle (a larva, a rare species, an obscure place), retrieval beats recall.
• Rights-aware pipelines — filter on rights.commercial and carry rights.attribution through to the final artifact automatically.

​

Direct REST use

Agents that don’t speak MCP can call the REST API directly. Search uses Voyage image-and-caption embeddings and text reranking. Inputs are text, image uploads, public HTTPS image URLs, or an image plus text. Use an indexed asset ID for similar-image lookup. See multimodal search for the request fields.