Add bounded query-page opportunities to GSC analytics and normalization

[AJaySi]

Motivation

• Provide query→page opportunity data (query, page, clicks, impressions, ctr, position) to downstream content-planning and agent workflows so they can make refresh-vs-new brief recommendations.
• Prevent oversized Search Console query+page payloads by bounding the request size and date window to keep API usage predictable and responses manageable.

Description

• Added constants QUERY_PAGE_OPPORTUNITIES_ROW_LIMIT = 2500 and QUERY_PAGE_OPPORTUNITIES_MAX_WINDOW_DAYS = 90 and use them for the GSC ['query','page'] request to bound rows and date window in backend/services/gsc_service.py.
• Implemented _get_query_page_opportunity_window to derive a bounded start/end date for the query+page request and included query_page_data.requested_window metadata in the returned analytics payload.
• Added query_page_opportunities processing in backend/services/analytics/handlers/gsc_handler.py to transform query+page rows into a prioritized list (fields: query, page, clicks, impressions, ctr, position) sorted by opportunity and capped (top 100) for downstream consumption, and included this field in both success and error/partial responses.
• Updated the autofill normalizer backend/api/content_planning/services/content_strategy/autofill/normalizers/analytics_normalizer.py to pass through query_page_opportunities into the normalized analytics structure.

Testing

• Compiled the modified modules with python -m compileall backend/services/gsc_service.py backend/services/analytics/handlers/gsc_handler.py backend/api/content_planning/services/content_strategy/autofill/normalizers/analytics_normalizer.py, which completed successfully.
• No additional automated unit tests were added or run as part of this change.

---

Codex Task