## [1] Instruction: architect-memory-mgmt

=== BEGIN instruction: architect-memory-mgmt ===

Before architecting your solution, consult these docs at the project root. Only read files that are likely to help (e.g., ignore `LOOK_AND_FEEL.md` for a memory leak issue):

- `./llmdocs/FILES.md`: Source file structure.
- `./llmdocs/FEATURES.md`: Implemented features.
- `./llmdocs/STACK.md`: Tech stack.
- `./llmdocs/LOOK_AND_FEEL.md`: Visual and UX design.
- `./llmdocs/IMPORTANT_NOTES.md`: Key implementation or project constraints.

Use this context to guide your architectural decisions.

Before returning your final results to your parent task, if you received XMl `doc_changes` sections from your child tasks' responses, then compile a detailed summary of all those `doc_changes` sections. This should include:

* File modifications (if any)
* Feature updates (if any)
* UI/UX changes (if any)
* Stack or architecture changes (if any)
* Key insights or lessons learned that could improve future decisions or avoid past mistakes (if any)

Wrap this summary between an XMl `doc_changes` tag and embed it directly into your response to the original (parent) task.

=== END instruction: architect-memory-mgmt ===

---INSTRUCTION-END---