Skip to main content

Seeding data

start the sandbox with a predictable dataset for repeatable tests

Seeding data

Seed files let you start the sandbox with predictable initial state, which is useful for:

  • reproducible integration tests
  • demos and tutorials
  • validating migrations and edge cases

Pass seeds at startup:

r1-plugins-sandbox --kv-seed ./seeds/cstore.json --fs-seed ./seeds/r1fs.json

CStore seed format​

--kv-seed expects a JSON array of entries with key and JSON value:

[
  {
    "key": "jobs:123",
    "value": { "status": "queued" }
  }
]

R1FS seed format​

--fs-seed expects a JSON array of entries with:

  • path (string)
  • base64 (string, base64-encoded file content)
  • optional content_type (string)
  • optional metadata (object)
  • optional last_modified (RFC3339 string)

Example:

[
  {
    "path": "/artifacts/report.json",
    "base64": "eyJvayI6IHRydWV9",
    "content_type": "application/json",
    "metadata": { "workflow": "ci" }
  }
]

Tips​

  • Keep seed files small and focused; use separate fixtures per test suite.
  • For binary payloads, base64-encode the raw bytes.

Next steps​