Final summary: • I answered your question only; I did not make any code or repository changes. • Main point: fixing failing CI snapshot tests is time-consuming because the agent must determine whether the snapshot change is intentional or a real regression, reproduce the CI environment accurately, inspect often-large and low-signal diffs, and run multiple verification loops. • Snapshot failures are especially slow because they are environment-sensitive and often indicate an underlying stability issue rather than a simple baseline update. • In short: the expensive part is not regenerating the snapshot, but proving the new snapshot is correct and stable.

Final summary: • I did not make any code or file changes. • I only answered your questions about snapshot-related CI failures and how to improve first-pass snapshot success on PRs. What I explained: • Snapshot fixes are time-consuming because the hard part is usually validating whether the snapshot should change, not regenerating it. • They are often slow due to: ◦ environment differences between local and CI, ◦ large/noisy diffs, ◦ nondeterministic inputs, ◦ multiple rerun cycles, ◦ and the risk of masking a real regression by blindly updating snapshots. What I recommended for getting future snapshot PRs to pass the first time: • Run the exact CI snapshot command locally before pushing. • Match the same environment as CI as closely as possible. • Make tests deterministic by fixing: ◦ timezone, ◦ locale, ◦ random seeds, ◦ generated IDs, ◦ sort order, ◦ test data. • Eliminate visual instability by: ◦ disabling animations, ◦ waiting for fonts/data/network to settle, ◦ using fixed viewports. • Keep snapshots small and targeted. • Only update snapshots after confirming the change is intentional. Status: • Your request was completely addressed. • No repository changes were made. • No extraneous changes exist to revert.