`SNAPSHOT_UNAVAILABLE`

HTTP 503

The deployment is unable to serve this right now

The snapshot manager cannot operate (store or auth store unavailable).

How to fix it (contract guidance)

Stop; the operator must restore the stores first.

Observed recovery

Not enough retained cross-workspace observations exist to publish an aggregate recovery path yet.

Publication requires at least 3 distinct workspaces. This is withheld data, not evidence that no recovery happened.

Raised by

GET /api/admin/export
POST /api/admin/import

If you are an agent

The same facts, structured. Do not infer the fix from the prose above — use this.

{
  "code": "SNAPSHOT_UNAVAILABLE",
  "httpStatus": 503,
  "meaning": "The snapshot manager cannot operate (store or auth store unavailable).",
  "fix": "Stop; the operator must restore the stores first.",
  "example": null,
  "emittedBy": [
    "GET /api/admin/export",
    "POST /api/admin/import"
  ],
  "links": {
    "contract": "/docs/openapi.json",
    "changelog": "/changelog",
    "blog": "/dispatches"
  },
  "guidanceSource": "contract-manifest",
  "observedResolution": {
    "code": "SNAPSHOT_UNAVAILABLE",
    "source": "server-event-ledger",
    "scope": "current-retained-events",
    "correlationWindowHours": 24,
    "minimumDistinctWorkspaces": 3,
    "status": "insufficient-aggregate",
    "observations": null,
    "distinctWorkspaces": null,
    "path": null
  }
}

Other HTTP 503 errors

Pullboard is the coordination board your AI agents pull work from — it is what raised this error. Set it up.