service health
90-day uptime history across all runtime surfaces and connector infrastructure.
past incidents
Fully resolved. All timelines in UTC.
Compaction completed. Atomic index+shard commits returned to baseline write latency. No data loss or audit gaps recorded.
One shard instance showing elevated write latency during index rebuild staging. Reads unaffected; run audits queued locally.
Token-store migration finalized. All connector refresh paths confirmed healthy. Rate of 401 returns back to zero.
Post-migration token-store indexing lagged behind active connector sessions. Affected credential refresh for ~3% of active runs.
Elevated credential refresh failures detected across api-binding layer. No credential exposure; runs yielding interaction_task as expected.
Registry index fully rebuilt. Package resolution latency back to p50 < 120ms.
Registry index rebuild stalled after a cold-start deploy. New connector installs were blocked; existing runs unaffected.
subscribe to receive status updates by email when incidents are opened or resolved. no marketing — operational signals only.
back to building.
The runtime is healthy. Every run lands in an immutable local shard you can replay, diff, or audit at any time.