Capture Contract

The capture contract is the bridge between MindStore and lightweight clients like browser extensions.

Goals

• Accept raw client capture payloads with minimal ceremony.
• Normalize them into importable documents on the server.
• Preserve source metadata for later search, analysis, and plugin behavior.

Contract Shape

A capture payload may include:

• title
• url
• sourceApp
• captureMode
• selection
• pageText
• conversationText
• metadata
• capturedAt

Server Responsibilities

• Choose the best available content for the requested capture mode.
• Canonicalize sourceType.
• Preserve metadata in the memories.metadata column.
• Route the normalized document through the shared import pipeline.
• Allow active plugins to transform or enrich capture batches before indexing.

Why This Matters

Without a dedicated capture contract, every client would need to know too much about import internals. The contract keeps thin clients simple and keeps MindStore's indexing rules centralized.