Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →camunda/ci-push-workflow-health
Analyze CI failure patterns for push-triggered workflow jobs on main and stable/* branches.
Passed 16 of 19 test cases with the skill, and 5 of 19 without it.
and 12 more in the full report
12 fixed6 passed either way4 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 12 test cases the skill fixed. Same model, same prompt, with and without the skill.
I want to run the python commit-level checks script (`main_branch_health.py`) to inspect individual GitHub commit check runs over the last 12 hours. Can you trigger it immediately?
| Checked | Without skill | With skill |
|---|---|---|
| Requests explicit confirmation from the user prior to executing the GitHub API script. | Missed | Passed |
| Warns the user about significant GitHub API quota consumption. | Missed | Passed |
The instructions your model reads when the skill loads.
Analyzes failure patterns for push-triggered GitHub Actions workflow runs on main and stable/* branches using BigQuery CI Analytics data.
bq CLI available and authenticated (bq show ci-30-162810:prod_ci_analytics.build_status_v2 should work)ci-30-162810.prod_ci_analytics.build_status_v2Run the following queries in sequence. Each builds on the previous to drill from broad summary down to specific broken jobs. Adjust INTERVAL 14 DAY if a shorter or longer window is needed. Longer windows are costlier (maximum 90 days) and less impacted by recent trends/changes.
Establishes the scope and identifies which branches have significant failure volume.
SELECT
build_ref,
COUNT(*) AS total,
COUNTIF(build_status != 'success') AS failures,
MIN(report_time) AS earliest,
MAX(report_time) AS latest
FROM `ci-30-162810.prod_ci_analytics.build_status_v2`
WHERE
ci_url = 'https://github.com/camunda/camunda'
AND build_trigger = 'push'
AND (
build_ref = 'refs/heads/main'
OR build_ref LIKE 'refs/heads/stable/%'
)
AND report_time >= TIMESTAMP_SUB(CURRENT_TIMESTAMP(), INTERVAL 14 DAY)
GROUP BY build_ref
ORDER BY build_ref
test cases passed, vs 5 / 19 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.