Background
PR #20 added six new cloud routines plus extensive cross-cutting controls. Several routines (Inspector on soul-day, Archivist on a covered estate, Apothecary on a clean alert day, etc.) will produce no-op runs by design. Over weeks, no-op runs are indistinguishable from silent failure (PAT expired, gist corrupted, predicate regex broken).
Request
Update routines/weekly-scorecard.prompt.md to add a new section:
Routines with zero actions this week
For each of the 12 cloud routines, read the most recent entry in its <routine>-state gist's run_log field. Report:
- Routine name and basename.
- Date of last action (any
action != no_op or skipped).
- If >7 days since last action: flag in Slack-readable format with the gist URL for inspection.
This catches silent failure within 7 days at no extra routine-runtime cost.
Why now
Per architecture-critic review of PR #20: "A routine producing only no-ops is indistinguishable from a routine that has silently broken." Weekly Scorecard is the natural place for this monitoring.
Provenance
Background
PR #20 added six new cloud routines plus extensive cross-cutting controls. Several routines (Inspector on
soul-day, Archivist on a covered estate, Apothecary on a clean alert day, etc.) will produce no-op runs by design. Over weeks, no-op runs are indistinguishable from silent failure (PAT expired, gist corrupted, predicate regex broken).Request
Update
routines/weekly-scorecard.prompt.mdto add a new section:For each of the 12 cloud routines, read the most recent entry in its
<routine>-stategist'srun_logfield. Report:action!=no_oporskipped).This catches silent failure within 7 days at no extra routine-runtime cost.
Why now
Per architecture-critic review of PR #20: "A routine producing only no-ops is indistinguishable from a routine that has silently broken." Weekly Scorecard is the natural place for this monitoring.
Provenance
cloud-routine