Set up a nightly code audit

Wake up to a fresh report on your codebase every morning — DRY/SOLID issues, security patches, and version bumps — produced by a pipeline of AI agents while you sleep. This is exactly how we keep our own codebase healthy.

What you'll use

The Orchestration engine in AI Studio — a multi-stage pipeline of agents that analyze your repo and write up findings — plus a schedule so it runs itself.

Steps

1. Open AI Studio → Operate → Orchestrations and click Start.
2. Pick the Nightly Codebase Health Pipeline from the catalog. It chains an analysis stage (scan for DRY/SOLID, security, dependency drift) into a report stage.
3. Point it at your repository: fill the run inputs (repo, branch). A SiloLink coding session does the actual scan on your machine; the agents reason over what it finds.
4. Set a cost limit so a run can never overspend, then launch. Watch the stages advance live and read the findings report it produces as an artifact.
5. Happy with a one-off run? Make it nightly: give it a schedule with a cron like 0 7 * * * (7am daily). The Scheduled Orchestration Launcher fires it automatically — no one has to press Start.

Result

Every morning there's a dated codebase-health report waiting, with prioritized, human-readable findings. Gate it with an approval step if you want a person to review before any fix work is dispatched.

Tip: the same shape works for a nightly dependency-upgrade or security-patch sweep — just pick the matching pipeline and schedule.