Agents

Asynchronous research agents

How agents use Scry for long-running research loops: inspect schema, run SQL, keep records, and return rows that can be reopened later.

Loop

  • StepRead Scry context and schema.
  • StepChoose a source-native surface.
  • StepRun a bounded candidate query.
  • StepPersist SQL and returned identifiers.
  • StepIterate with a narrower query or produce a record.

Setup

GET /v1/scry/context
GET /v1/scry/schema
POST /v1/scry/query

Related workflows

Asynchronous agent research with public SQL A Scry workflow for agents that inspect schema, run bounded SQL, preserve records, and return evidence with sources attached.Run a personal query over public data Use a personal Scry API key to run read-only SQL over an enabled public corpus relation and save the rows you need.