Broken pipelines stay silent until someone notices a dashboard looks wrong, usually days later. This watcher catches the break when it happens.
sonnet4 daysdbtPostgresGreat ExpectationsSlack
Claude
60ROI
72Scale
$2.8k93Saved
ROI for
README.md
Why this subagent
Broken pipelines stay silent until someone notices a dashboard looks wrong, usually days later. This watcher catches the break when it happens.
It profiles each table on a schedule for freshness, row counts, and schema, then compares against the last known-good run. A drop, a stale load, or a changed column triggers a Slack alert to the owning team, naming the table, the failed check, and the last value that looked right.
How it runs
Used at step 01 to kick off the pipeline.
Write
Used at step 01 to kick off the pipeline.
WebFetch
Used at step 01 to kick off the pipeline.
WebSearch
Used at step 01 to kick off the pipeline.
Compare every run against the prior baseline to catch drops, stale loads, and added or dropped columns.
pending
Classify each issue by severity based on how far it deviates from the table's normal range.
pending
Alert the owning team in Slack with the table, the check that failed, and the last good value.
pending
Sample output
json
// Sample output
// (generated when the pipeline finishes)
On a schedule, check tracked tables for freshness, row counts, and schema drift, and alert owners on any deviation.
Unlock the rest
The full agent definition, install snippet, and starter task are gated for community members.
Members get the full `.md` agent file, the npm / pnpm install one-liners, a starter prompt that we've tuned against real runs, and the open-source repo when this automation ships there. One email, magic link, done.