log(ger)

A personal productivity tracking web app — because every minute is worth logging.

---

A ground-up rebuild of Chatable-Study-Database (Python/Jupyter) as a modern web application. Tracks study and work time across academic sessions with real-time timers, analytics dashboards, AI-powered chat, and project progression visualization.

Features

• Timer-based logging — Start/pause/stop timers per category with parallel sessions
• Manual entry — Retroactive time logging for sessions you forgot to time
• Analytics dashboard — Time-scale-aware filtering (Overall / Year / Month / Week) with stacked area charts, category breakdowns, session comparison by family groups, and weekly heatmaps
• AI chat — Claude API-powered conversational interface to query past activity and generate summaries
• CSV import — Import legacy data with two-panel upload (timesheet CSV + text/notes CSV). 14 sessions from Fall 2022 through Winter 2026
• Category families & groups — Two-level hierarchy: groups (Research, Courses, Personal) → families (Salk Research, COGS, Training) → categories. Full family management UI (create, edit, rename, delete) with inline editing on the Data page
• Project progression — Bubble visualization of how project families evolve across sessions
• Smart normalization — CSV columns are clean display names; session info comes from filenames, not column name suffixes
• Auto-active sessions — Imported sessions whose date range includes today are automatically marked active

Tech Stack

Layer	Technology
Frontend	SvelteKit 2, Svelte 5 (runes), Tailwind CSS 4, D3.js, TypeScript
Backend	FastAPI, SQLAlchemy 2 (async), Pydantic 2
Database	SQLite (local)
AI	Claude API (Anthropic Python SDK)
Package managers	uv (Python), pnpm (Node)

Quick Start

Prerequisites

• Python 3.11+
• Node.js 22+
• uv (Python package manager)
• pnpm (Node package manager)

Install

# Backend
cd backend
uv sync

# Frontend
cd frontend
pnpm install

Import Legacy Data

If you have CSV data in data/, run the batch import:

# Start backend first
cd backend
uv run uvicorn logger.main:app --port 8000

# In another terminal, trigger batch import
curl -X POST http://localhost:8000/api/import/batch

CSV files follow the naming convention {year}_{season}_study.csv and {year}_{season}_text.csv (e.g., 2024_fall_study.csv). Category columns should be clean display names — session context comes from the filename.

Run

# Terminal 1 — Backend (port 8000)
cd backend
uv run uvicorn logger.main:app --reload --port 8000

# Terminal 2 — Frontend (port 5173)
cd frontend
pnpm dev

Open http://localhost:5173.

Project Structure

logger/
├── backend/
│   ├── pyproject.toml
│   ├── import_all.py              # Batch CSV import script
│   └── logger/
│       ├── main.py                # FastAPI app + CORS + lifespan
│       ├── config.py              # Paths, DB URL, CORS origins
│       ├── database.py            # Async SQLAlchemy engine, init_db()
│       ├── models.py              # ORM models (Session, Category, etc.)
│       ├── schemas.py             # Pydantic request/response schemas
│       ├── routers/
│       │   ├── sessions.py        # CRUD + active session
│       │   ├── categories.py      # Categories + families CRUD
│       │   ├── analytics.py       # Overview, daily series, categories, heatmap
│       │   ├── daily.py           # Daily activity + streak
│       │   ├── timers.py          # Timer start/pause/resume/stop
│       │   ├── manual_entries.py  # Manual time entries
│       │   ├── import_csv.py      # CSV preview + confirm + batch
│       │   ├── chat.py            # AI chat endpoints
│       │   └── settings.py        # Key-value settings + DB info
│       ├── services/
│       │   ├── import_service.py       # CSV parsing + DB writes
│       │   ├── family_service.py       # Family auto-detection
│       │   ├── category_normalization.py # Clean name → merge key
│       │   ├── analytics_service.py    # Analytics queries
│       │   ├── timer_service.py        # Timer logic
│       │   └── chat_*.py              # Chat context + query services
│       └── utils/
│           ├── csv_utils.py        # BOM handling, header parsing
│           └── date_utils.py       # Date parsing, day normalization
├── frontend/
│   ├── package.json
│   ├── vite.config.ts
│   └── src/
│       ├── app.css                 # Tailwind + theme variables
│       ├── app.html
│       ├── lib/
│       │   ├── api/client.ts       # Typed API client
│       │   ├── stores/             # Session, timer, theme, chat stores
│       │   ├── assets/             # Logo SVGs
│       │   └── components/
│       │       ├── layout/         # Sidebar, TopBar
│       │       ├── dashboard/      # StatCard, ActiveTimers
│       │       ├── timer/          # TimerCard, StopDialog, ManualEntryForm, TodayLog
│       │       ├── analytics/      # FilterBar, DailyAreaChart, CategoryBars, SessionBars, WeeklyHeatmap
│       │       ├── data/           # SessionList, SessionDetail, ImportDropzone, BubbleVisualization, FamilyManager
│       │       └── chat/           # ChatMessage, ChatInput
│       └── routes/
│           ├── +layout.svelte      # App shell
│           ├── +page.svelte        # Dashboard
│           ├── timer/              # Timer + manual entry
│           ├── analytics/          # Analytics with time-scale filtering
│           ├── chat/               # AI chat interface
│           ├── projects/           # Project progression
│           ├── data/               # Session list + CSV import + family manager (3 tabs)
│           └── settings/           # Theme, DB info, API key
├── data/                           # CSV files (gitignored)
├── scripts/                        # Data cleaning scripts
└── logger.db                       # SQLite database (gitignored)

Data Model

• sessions — One per academic term (e.g., Winter 2026)
• category_families — Links related categories across sessions (e.g., "Salk Research"), with family_type for group-level classification (research/course/personal)
• categories — Individual tracked activities within a session, linked to a family
• daily_records — One row per date per session with total minutes and week number
• observations — Time-per-category data (long format, one row per day/category pair)
• text_entries — Free-text daily descriptions (location, notes, study materials)
• timer_entries — Real-time timer sessions with pause/resume support
• manual_entries — Retroactive time entries
• category_groups — Auto-generated groupings for visualization
• ai_descriptions / chat_messages — AI-generated content and chat history
• v_daily_totals / v_family_totals — Materialized views for analytics

Implementation Status

• Foundation — Backend, database, CSV import, app shell, data page
• Timer & Manual Entry — Real-time timers, manual logging, dashboard with today's log
• Analytics — Time-scale filtering, stacked area chart, category breakdown with session labels, session comparison by family groups, weekly heatmap
• Data Management — Two-panel CSV import, full family CRUD (create/edit/rename/delete), bubble visualization, three-tab data page (Sessions, Groups, Families)
• CSV Normalization — Clean column names in CSVs, simplified import pipeline, family auto-detection from display names, auto-active session detection on import
• AI Chat — Claude API integration with approval flow
• Projects — Family timeline with AI-generated descriptions
• Polish — Responsive layout, loading states, error handling

License

Private project.