[Chore] Refactor shared/api_utils.py to use penguin-libs

[PenguinzTech]

User Story

As a maintainer, I want to refactor shared/api_utils.py to leverage penguin-libs utilities so that common API patterns use the shared library instead of local implementations.

Background

shared/api_utils.py provides ~150 lines of generic REST API helpers:

• paginate(query, page, per_page) — SQLAlchemy query pagination
• get_pagination_params() — Extract page/per_page from request args
• validate_request(schema_class, data) — Marshmallow validation
• make_error_response(message, status_code) — Error response formatting
• make_success_response(data, message, status_code) — Success response formatting
• apply_filters(query, model, filters) — Query filtering
• get_or_404(model, id) — Fetch or abort

Elder already uses penguin_libs.pydantic.flask_integration.validated_request for request validation in 12+ endpoint files, partially overlapping with validate_request().

Acceptance Criteria

• Audit which api_utils functions are still used vs replaced by penguin-libs
• Remove functions superseded by penguin-libs (e.g., validate_request → validated_request)
• Keep Elder-specific helpers that have no penguin-libs equivalent
• If penguin-libs gains pagination/response utils, migrate to those
• All API endpoints continue working
• Tests pass (unit + integration)
• Linting passes

Notes

• Low effort — mostly removing dead code and consolidating imports
• Some functions may need to stay as Elder-specific thin wrappers
• Consider contributing generic pagination/response utilities back to penguin-libs