# Why production might not match main

Prompt · Infrastructure and environment · I02

Lists every way 'merged' can fail to mean 'live'.

## Prompt

```text
Trace a change from commit to production, including builds, migrations, deployment triggers, cache refreshes, and background services. Identify every reason production might not reflect merged code.
```

**What it returns:** The real pipeline as steps with their config files, and for each, how it can skip or fail without anyone noticing.

**Skip it when:** Right after the same trace, unless the release process changed.

**Tools:** Automated build and Vercel config, deployment history (read-only).

**Changes:** Changes nothing

**Sessions:** One session

## Related

- [Database changes during a deploy](https://atlasscope.dev/prompts/database-changes-during-a-deploy): Checks that each database change works with the old code that is still live while a release finishes.
- [Dashboard and repo out of sync](https://atlasscope.dev/prompts/dashboard-and-repo-out-of-sync): Finds where the hosting dashboard and the repo disagree.
- [Can we rebuild this database from its sources?](https://atlasscope.dev/prompts/can-we-rebuild-this-database-from-its-sources): You worry about disaster recovery, production-only state, or whether the data could be rebuilt.

Source: https://atlasscope.dev/prompts/why-production-might-not-match-main