Audits and budgets
Run the audit once, then wire it to a budget the pipeline enforces on every pull request.
An audit that ends in a report changes nothing. People read it once and fix a few things, then the same problems come back next quarter.
Run it once and fix what it finds, then set the limit in the pipeline as a budget. After that the problem cannot come back unnoticed.
Close the metadata and crawlability findings for one public surface
Each finding fixed and the audit rerun clean
Find and fix broken links and missing files
Every link resolving, the check running on a schedule afterwards
Run the performance audit on one key route and land its top finding
Score before and after, with the audit output attached
Measure unused CSS and JavaScript on one route
Unused bytes reported before and after
Reconcile field measurements against the lab audit
Both sources reported side by side in the pull request
Turn a framework build warning category into an error
The category reads empty and the build fails if it returns
Add the accessibility audit to the pipeline
A seeded violation fails the build
Put the dependency and license audit on a schedule
A scheduled run opens work the day an advisory lands
Set a bundle-size budget per entry point
The budget fails on a seeded import
Wire the performance audit into the pipeline with a budget per route
The budget fails on a deliberately regressed route
Put the agent team to work on your backlog
Connect a repository and assign SHIP your first issue. Your first missions are free.