Added -Natural Language Command Routing and more -v1.3.0

[1.3.0] — 2026-03-24

✨ New Features

Natural Language Command Routing (notionmind.py)

• Type anything in plain English — Groq routes it to the right command automatically
• "who is Sachin Tendulkar" → routes to ask
• "i studied for JEE maths today" → routes to save
• "show me my notes" → routes to list
• Exact command names bypass Groq entirely — zero latency for power users
• Unknown inputs handled gracefully with a helpful message
• show_help() — clean grouped command panel replacing the messy choices list

Todo List (todos.py, notionmind.py, analytics.py)

• New todos.py — full todo list stored in Notion database
• Add todos with priority (high/medium/low), due date, and optional note
• Priority stored as priority-high/medium/low tags in Notion
• List pending todos — sorted by priority then due date
• Mark complete — moves from todo tag to todo-done tag
• Delete todos with confirmation
• View completed todos
• Overdue todos highlighted in red, today's todos in yellow
• todo command in interactive mode and CLI
• Todos panel in dashboard — shows up to 6 pending todos with priority icons

Analytics Dashboard (analytics.py, notionmind.py)

• New analytics.py — stunning terminal analytics dashboard
• dashboard command — full panel layout:
  • Row 1: GitHub-style 30-day activity heatmap + quick stats panel
  • Row 2: 14-day knowledge growth bar chart + topic velocity with trend arrows
  • Row 3: Upcoming reminders panel + pending todos panel (side by side)
• Activity heatmap uses heat characters (· ▪ ▫ ▬ █) colored grey → bright green
• Topic velocity shows each tag's count this week vs last week with ↑ ↓ → trend arrows
• insights command — AI-powered personal insights using Groq llama-3.3-70b:
  • 🔥 Peak Productivity — when you're most active, specific dates cited
  • 🕳 Knowledge Gaps — topics being underexplored given current work
  • 📉 Fading Topics — what you explored before but drifted from
  • ⚡ This Week's Action — one sharp actionable recommendation
  • Each section rendered as a distinct colored Rich panel

Semantic Search (brain.py)

• Vector similarity search using fastembed locally
• Model: BAAI/bge-small-en-v1.5 (~67MB, cached at ~/.cache/fastembed)
• Runs entirely on CPU — no API, no cost, no internet after first download
• Fetches full Notion page block content — not just summary
• Notes cache — fetches all notes + full content once per session
• Filters out auto-generated, summary, daily, weekly-report, category, merged notes
• Threshold: 0.55 cosine similarity minimum
• Attribute-aware: "cricketer from Kerala" correctly finds the Kerala cricketer note
• 7. Search in graph menu

Think — Multi-hop Reasoning (brain.py)

• Complex question answering across notes + graph
• Extracts keywords using Groq, searches each separately
• Expands one hop via knowledge graph edges
• Strict prompt — never invents note titles
• 8. Think in graph menu

Recall — Knowledge Evolution (brain.py)

• Chronological topic understanding analysis
• Finds all notes semantically related to a topic, sorted by date
• Groq narrates: initial curiosity → experiments → insights → current depth
• 9. Recall in graph menu

Suggest — Auto Related Notes (brain.py, notionmind.py)

• Groq extracts the single main keyword from a saved note
• Ignores noise words like "today", "watched", "done"
• Searches title + summary only — fast, no full page fetch
• Keyword split matching — "jee maths" matches notes with "jee" or "maths"
• Shows up to 3 related notes after every save

Full-featured Telegram Bot (telegram_bot.py)

• All commands available on Telegram:
• /save, /ask, /today, /list, /search, /semantic, /read, /delete
• /stats, /export, /inbox, /results, /weekly
• /todos, /addtodo, /donetodo
• /remind, /reminders
• /think, /recall, /dashboard, /insights
• /menu — shows all available commands in a grouped list

---