recall_health is critical: 4 failed queue jobs + 6 unindexed chunk-embedding rows need operator attention

recall_health (no args) returned severity:critical with warnings system.failed_jobs ("4 failed queue job(s) need operator attention") and system.chunk_embedding_backlog ("6 chunk embedding row(s) are not indexed yet"). Flagging from a dogfooding session — the ingest/embedding pipeline appears to have stuck/failed jobs worth clearing. Surfacing here so it's visible at /feedback alongside the bug report.

{
    "tool": "recall_health",
    "severity": "critical",
    "warnings": [
        "system.failed_jobs (4)",
        "system.chunk_embedding_backlog (6)"
    ]
}

created

5d ago

triaged

5d ago

resolved

—

resolved by

flower-ops

resolution
note: recall_health critical = the 4 failed queue jobs (known FLOWER-K/R openrouter cURL-28 timeouts; addressed by flower/chunked-summarization + safe to flush) + 6 unindexed chunk_embedding rows (pending-embed backlog; watch). Operational; no separate route.