6 Commits

Author	SHA1	Message	Date
Lars	1298bd235f	feat: add structured JSON logging for all feature usage (Phase 2) ... - Create feature_logger.py with JSON logging infrastructure - Add log_feature_usage() calls to all 9 routers after check_feature_access() - Logs written to /app/logs/feature-usage.log - Tracks all usage (not just violations) for future analysis - Phase 2: Non-blocking monitoring complete Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-20 22:18:12 +01:00
Lars	ddcd2f4350	feat: v9c Phase 2 - Backend Non-Blocking Logging (12 Endpoints) ... PHASE 2: Backend Non-Blocking Logging - KOMPLETT Instrumentierte Endpoints (12): - Data: weight, circumference, caliper, nutrition, activity, photos (6) - AI: insights/run/{slug}, insights/pipeline (2) - Export: csv, json, zip (3) - Import: zip (1) Pattern implementiert: - check_feature_access() VOR Operation (non-blocking) - [FEATURE-LIMIT] Logging wenn Limit überschritten - increment_feature_usage() NACH Operation - Alte Permission-Checks bleiben aktiv Features geprüft: - weight_entries, circumference_entries, caliper_entries - nutrition_entries, activity_entries, photos - ai_calls, ai_pipeline - data_export, data_import Monitoring: 1-2 Wochen Log-Only-Phase Logs zeigen: Wie oft würde blockiert werden? Nächste Phase: Frontend Display (Usage-Counter) Phase 1 (Cleanup) + Phase 2 (Logging) vollständig! Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-20 21:59:33 +01:00
Lars	e4f49c0351	fix: enable AI analysis history and correct pipeline scope ... Fixes two critical bugs in AI analysis storage: 1. History now works - analyses are saved, not overwritten - Removed DELETE statements before INSERT in insights.py - All analyses are now preserved per scope - Displayed in descending order by creation date 2. Pipeline saves under correct scope 'pipeline' instead of 'gesamt' - Changed scope from 'gesamt' to 'pipeline' in pipeline endpoint - Pipeline results now appear under correct category in history 3. Fixed pipeline appearing twice in UI - Filter now excludes both 'pipeline_*' and 'pipeline' from individual list - Pipeline only appears in dedicated section at top Changes: - backend/routers/insights.py: Removed DELETE, changed scope to 'pipeline' - frontend/src/pages/Analysis.jsx: Fixed filter to exclude 'pipeline' Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-20 15:35:33 +01:00
Lars	4fcde4abfb	ROLLBACK: complete removal of broken feature enforcement system ... Reverts all feature enforcement changes (commits 3745ebd, cbad50a, cd4d912, 8415509) to restore original working functionality. Issues caused by feature enforcement implementation: - Export buttons disappeared and never reappeared - KI analysis counter not incrementing - New analyses not saving - Pipeline appearing twice - Many core features broken Restored files to working state before enforcement implementation (commit 0210844): - Backend: auth.py, insights.py, exportdata.py, importdata.py, nutrition.py, activity.py - Frontend: Analysis.jsx, SettingsPage.jsx, api.js - Removed: FeatureGate.jsx, useFeatureAccess.js The original simple AI limit system (ai_enabled, ai_limit_day) is now active again. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-20 15:19:56 +01:00
Lars	3745ebd6cd	feat: implement v9c feature enforcement system ... Backend: - Add feature access checks to insights, export, import endpoints - Enforce ai_calls, ai_pipeline, data_export, csv_import limits - Return HTTP 403 (disabled) or 429 (limit exceeded) Frontend: - Create useFeatureAccess hook for feature checking - Create FeatureGate/FeatureBadge components - Gate KI-Analysen in Analysis page - Gate Export/Import in Settings page - Show usage counters (e.g. "3/10") Docs: - Update CLAUDE.md with implementation status Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-20 12:43:41 +01:00
Lars	b4a1856f79	refactor: modular backend architecture with 14 router modules ... Phase 2 Complete - Backend Refactoring: - Extracted all endpoints to dedicated router modules - main.py: 1878 → 75 lines (-96% reduction) - Created modular structure for maintainability Router Structure (60 endpoints total): ├── auth.py - 7 endpoints (login, logout, password reset) ├── profiles.py - 7 endpoints (CRUD + current user) ├── weight.py - 5 endpoints (tracking + stats) ├── circumference.py - 4 endpoints (body measurements) ├── caliper.py - 4 endpoints (skinfold tracking) ├── activity.py - 6 endpoints (workouts + Apple Health import) ├── nutrition.py - 4 endpoints (diet + FDDB import) ├── photos.py - 3 endpoints (progress photos) ├── insights.py - 8 endpoints (AI analysis + pipeline) ├── prompts.py - 2 endpoints (AI prompt management) ├── admin.py - 7 endpoints (user management) ├── stats.py - 1 endpoint (dashboard stats) ├── exportdata.py - 3 endpoints (CSV/JSON/ZIP export) └── importdata.py - 1 endpoint (ZIP import) Core modules maintained: - db.py: PostgreSQL connection + helpers - auth.py: Auth functions (hash, verify, sessions) - models.py: 11 Pydantic models Benefits: - Self-contained modules with clear responsibilities - Easier to navigate and modify specific features - Improved code organization and readability - 100% functional compatibility maintained - All syntax checks passed Updated CLAUDE.md with new architecture documentation. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-19 11:15:35 +01:00