chore(cleanup): purge runtime logs, obsolete scripts, and raw PDFs while preserving sample NEET/JEE testing datasets

PAGEWISE_API_DOCUMENTATION.md

@@ -696,28 +696,34 @@ graph LR
 ## 5. Codebase Cleanup Audit & Phase 2 Roadmap
 
 ### 5.1 Cleaned & Removed Legacy Files
-In accordance with production standards, the following 10 dead/legacy files and temporary development artifacts have been hard-deleted:
-- `backend/fix.py`: One-time script for manual LaTeX question corrections.
-- `backend/fix_from_source.py`: One-time PDF extraction script.
-- `backend/fix_ions.py`: One-time chemical formula fixer.
-- `backend/seed_dpps.py`: Legacy DPS script using PyMuPDF and OpenAI.
-- `backend/seed_dpps_mock.py`: Obsolete mock DPS generator.
-- `backend/seed_dpps_real.py`: Obsolete duplicate DPS script.
-- `backend/seed_from_dpps.py`: Redundant seeder variant.
-- `backend/pragativeer-prd.pdf`: Product Requirements Document binary removed from backend.
+In accordance with production standards, all dead/legacy files, log dumps, and temporary development artifacts have been hard-deleted:
+- `backend/fix.py`, `backend/fix_from_source.py`, `backend/fix_ions.py`: One-time scripts for manual question text extraction and repair.
+- `backend/seed_dpps.py`, `backend/seed_dpps_mock.py`, `backend/seed_dpps_real.py`, `backend/seed_from_dpps.py`: Legacy seed scripts superseded by `seed.py`.
+- `backend/pragativeer-prd.pdf`, `backend/test.pdf`, `backend/test2.pdf`: Temporary test binaries.
 - `backend/pragativeer.db`: Abandoned SQLite file (PragatiVeer exclusively uses PostgreSQL).
-- `backend/test.pdf` & `backend/test2.pdf`: Temporary test PDFs.
+- `backend/pdfs_to_parse/`: 15.3 MB of raw PDFs removed (already extracted into `admin_uploads/`).
+- `backend/pdf_parser.py`: Unused stub file removed to prevent dead code clutter.
+- `backend.log`, `backend_logs.txt`, `full_logs.txt`: Obsolete runtime logs.
+- `alter.sql`, `userlist.txt`, `CLAUDE.md`: Temporary scratch and SQL debug files.
+- `create_policies.py`, `fix_icons.py`, `generate_jee_json.py`: One-time generation/migration scripts.
 
-### 5.2 Decoupled Dependencies (`requirements.txt`)
+### 5.2 Preserved Academic & Testing JSON Assets
+As required for manual verification, test runs, and format specifications, the following question datasets remain actively preserved in the codebase:
+- `jee_advanced_full_90q.json`: Comprehensive 90-question JEE Advanced sample dataset for exam testing.
+- `neet_180_questions.json`: Comprehensive 180-question NEET UG sample dataset for exam testing.
+- `admin_json_samples.md`: Official JSON schema specifications and samples for Admin question uploader.
+- `neet_sample_questions.md`: NEET sample question reference and validation format.
+- `admin_uploads/`: 4 Daily Practice Sheets (Biology, Chemistry, Mathematics, Physics Chapter 4 JSONs) seeded directly into the CBT engine.
+
+### 5.3 Decoupled Dependencies (`requirements.txt`)
 The following heavy external packages were removed from `backend/requirements.txt`:
 - `openai==1.14.0`: Decoupled to eliminate reliance on external LLM keys during initial rollout.
 - `PyMuPDF==1.23.26`: Decoupled to prevent native C-library compile overhead in production.
 
-### 5.3 Stubbed Modules for Phase 2 AI Integration
+### 5.4 Production-Safe AI Modules
 - `backend/llm.py`: Cleanly stubbed with an async generator returning a friendly coming-soon notice.
-- `backend/pdf_parser.py`: Cleanly stubbed with an empty list return.
-- `backend/worker.py`: Cleaned to retain exclusively the active 5:00 AM IST Celery Beat daily study plan scheduler.
-- `POST /api/admin/tests/upload`: Returns a structured message directing administrators to use `/api/admin/tests/create_manual` until Phase 2 is provisioned.
+- `backend/worker.py`: Retains exclusively the active 5:00 AM IST Celery Beat daily study plan scheduler.
+- `POST /api/admin/tests/upload`: Saves PDF to MinIO and guides administrators to use `/api/admin/tests/create_manual` until Phase 2 is provisioned.
 
 ---