calm-artist-7206
09/04/2026, 10:57 PMfresh-machine-37269
09/04/2026, 10:57 PMfresh-machine-37269
09/04/2026, 10:58 PMcalm-artist-7206
09/04/2026, 10:59 PMfresh-machine-37269
09/04/2026, 10:59 PM/var/lib/agent-canvas/.openhands/automation/automations.db in read-only mode (I didn't modify anything).
How I counted:
1. The automations table has 36 rows total.
2. 13 of those have a deleted_at timestamp set (soft-deleted), so I excluded them β they aren't "defined" as active Automations.
3. That leaves 23 active (non-deleted) Automations.
4. Of those 23, filtering on enabled = true gives 15 enabled and 8 disabled.
The breakdown:
So: 23 defined, 15 enabled.
One nuance worth flagging: if you meant "defined" to include everything ever created (including soft-deleted rows), the raw count is 36. I interpreted "defined" as active/current Automations, which is the 23. Happy to restate it either way., with interactive elementscalm-artist-7206
09/04/2026, 11:00 PMfresh-machine-37269
09/04/2026, 11:01 PMcalm-artist-7206
09/04/2026, 11:26 PMfresh-machine-37269
09/04/2026, 11:26 PMOpenHands/software-agent-sdk. Its prompt explicitly scopes it to that single repo ("Conduct a weekly regression hunt for `OpenHands/software-agent-sdk`").
Weekly Automation & Conversation Health Review β does not target any source repository. It's a read-only health/improvement audit of the shared OSS automation VM itself (dev-oss-automation-vm). Its data sources are local: the automations.db file, the dev_conversations/ directory, the KV store, and process list β not GitHub repos.
So: one repo (software-agent-sdk) for the regression hunt; no repos for the health review.