Status

Polls the per-query CloudWatch streams for a Retriever query that is still running, or that the prior Query call left as partialResults: true (poll budget exceeded). Returns a fresh diagnostics snapshot (scan stats, worker progress, errors) plus a verdict (complete, in-flight, scan pending, complete with zero events). Does not re-fetch result events from S3; rerun Query to pick those up once the verdict is complete.

Example

"status on retriever query b27c9-..."

Plan: 12 template hashes · 4 var sets · dispatch=remote · timeslice=60000ms

Scan: 826 scanned · 655 matched · 142 skipped (search) · 0 (template)

Workers: 36/120 complete · 41 MB fetched · 21,751 result events decoded

Status: in-flight — 36/120 workers complete. Check again in a few seconds.

More to ask

• "status 7e3-..., started at 1777200000"
• "is acme-corp-retry still scanning?"
• "why did b27c9-... return zero events?"

Prerequisites

This tool requires the Retriever deployed plus LOG10X_RETRIEVER_LOG_GROUP set to the per-query CloudWatch log group. When the env var is unset, the tool returns a polling_error rather than failing silently.

Tool schema (advanced)

Field	Type	Required	Default	Description
queryId	string	yes	—	The queryId returned by Query.
queryStartedAt	number	no	now − 5 min	Epoch milliseconds of when the original query was submitted. Bounds the CloudWatch scan to events from that point on.
environment	string	no	—	Environment nickname (multi-env).

Status verdicts:

• complete — workers finished and decoded events; rerun Query to retrieve them.
• complete, zero events — workers finished but matched nothing. The verdict line carries a classification reason (Bloom miss / false positive / stale indexer / empty search expression).
• in-flight — some workers still running. Re-poll in a few seconds.
• scan pending — coordinator dispatched but no scan stats yet. Either the scan is running or CloudWatch events have not flushed.
• unknown — no CloudWatch events for this queryId yet. Verify the queryId is correct.