api.honcho.dev timeouts impacting Hermes memory provider

[pax-g-bot]

Summary

api.honcho.dev is currently timing out from a Hermes Agent deployment in the UK, while the public status page appears to show all systems operational. This is outside the last maintenance window we were notified about.

This looks similar to the application/API timeout pattern previously reported in #753.

Timeline / context

• First watchdog alert observed: 2026-07-25 around 09:03 BST
• Rechecked: 2026-07-25 around 11:03 BST
• Last known forwarded maintenance window: 2026-07-18 07:00–10:00 New York time (12:00–15:00 BST), expected 20–30 minute interruption
• Current date is outside that planned window
• Public status page: appears operational from browser/search extraction, with no current incident visible

Observed behaviour

Hermes Honcho memory provider probe is failing repeatedly:

TimeoutError: Request timed out after 30.0s

Latest local watchdog state:

last_checked_at: 2026-07-25T10:03:52.022868+00:00
last_ok: False
last_detail: TimeoutError: Request timed out after 30.0s
last_elapsed_seconds: 92.247
last_status_page_detail: status page HTTP 403
maintenance_start_utc: 2026-07-18T11:00:00+00:00
maintenance_end_utc: 2026-07-18T14:00:00+00:00

Direct unauthenticated API probes from the same server also time out:

GET https://api.honcho.dev/health -> ReadTimeout after 15s
GET https://api.honcho.dev/       -> ReadTimeout after 15s

Impact

Hermes Agent is configured to use Honcho as its memory provider. While core chat and non-Honcho tools continue to work, Honcho-backed memory provider startup/recall/write paths are degraded or unavailable.

Expected behaviour

Either:

• API endpoints respond normally; or
• status page reflects a known degradation/incident if there is one.

Request

Could you confirm whether this is a known incident on the Honcho side, or if there is anything useful we should check from our side?

[themmm]

Same for me (Germany)