chore: ignore generated data directories in scaffolds

[steve8708]

Summary

• ignore every data/ directory in the framework checkout
• propagate the rule through first-party template and bundled scaffold _gitignore sources
• remove the tracked local PGlite lock from the repository while preserving local runtime data

Validation

• git diff --check
• scaffold source coverage check for all _gitignore files
• git check-ignore --no-index for root, nested, and generated-template data paths

The generated data directories are local runtime state and must not enter generated Agent-Native repos.

[builder-io-integration]

Builder reviewed your changes — looks good ✅

Review Details

Code Review Summary

PR #4610 broadens the repository and scaffold .gitignore rules from only selected runtime paths to every data/ directory, updates the first-party template sources consistently, and removes the stale tracked PGlite lock file. The approach is sound: the **/data/ pattern covers both root-level and nested generated runtime directories, while the existing runtime-specific entries remain harmless and preserve documentation of expected data contents. Independent reviews also confirmed that PGlite locking remains runtime-managed and that no other tracked data/ artifacts are left behind.

This is a low-risk repository/scaffold configuration change. No confirmed bugs, data-loss risks, crashes, or security issues were found. The source coverage and ignore semantics match the stated goal.

🧪 Browser testing: Skipped — PR only modifies backend/config/docs, no UI impact

[github-actions]

Here's a visual recap of what changed:

Open the full interactive recap