AazDoh (آزدوہ)

Keep the promises you make to yourself.
Commit • Do • Report • Reflect — Zero Dopamine Streaks, Pure Operational Integrity.

---

📑 Table of Contents

1. 📖 Overview & Philosophy
2. ⚡ Core Capabilities
3. 🏗️ Monorepo Architecture
4. 💻 Terminal CLI Agent (aazdoh-cli)
   • CLI Core Highlights
   • Installation & Setup
   • Zero-Friction Terminal Cockpit (az)
   • Complete CLI Command Reference
   • Interactive REPL Slash Commands
   • Configuration & Credential Hierarchy
   • Security & Input Guardrails
5. 🤖 Model Context Protocol (aazdoh-mcp)
   • Supported AI Environments
   • IDE Configuration Snippets
   • Complete Suite of 18 Native MCP Tools
   • AI Prompting Recipes & Workflows
   • Daily Developer Operational Playbook
6. 🚀 Quickstart & Installation
   • Prerequisites
   • 1. Database Setup
   • 2. Backend Configuration
   • 3. Frontend Setup
   • 4. CLI Setup
   • 5. MCP Server Setup
7. 🧪 Testing & Build Verification
8. 🎨 Design System (Kashmir Harud Aesthetic)
9. 📄 License

---

📖 Overview & Philosophy

Standard to-do apps reward dopamine checkmarks, artificial streaks, and passive backlog hoarding. AazDoh (Kashmiri: “Do it today”) is an Empirical Behavioral Operating System engineered for software engineers, founders, and focused operators.

Why AazDoh?

• No Dopamine Streaks: Streaks create fragile all-or-nothing psychology. AazDoh measures 52-week compounding consistency and cognitive stamina thresholds.
• Pre-Flight Overload Defense: Stop planning fallacy before your day begins by stress-testing proposed commitments against your empirical 7-day velocity.
• Confronting Rationalizations: When you postpone a commitment, AazDoh matches your justification against your historical excuse receipts.
• Developer Ecosystem Native: Work seamlessly across Web UI, autonomous Terminal CLI (aazdoh), or directly inside AI Coding IDEs (Claude Desktop, Cursor, Antigravity) via native MCP tools.

---

⚡ Core Capabilities

1. 🎯 Daily Commitments & Optional Day Phasing

• Active Focus First: Default views prioritize actionable commitments (Active Focus → Kept → All).
• 🌅 Day Phasing (Morning / Day / Evening / Anytime): Organize tasks into natural daily cognitive rhythms without brittle, anxiety-inducing hourly time-blocking.
• ✋ Drag & Drop Reordering: Seamlessly move commitment cards between day phases on the web UI with instant optimistic UI updates.
• Bounded Capacity: Encourages sustainable limits (2–4 high-leverage promises) rather than endless backlog anxiety.
• Dynamic Victory State: Visual celebration and telemetry locking once daily active promises are cleared.

2. 🛡️ 60-Second Plan Feasibility Check (Stress-Testing)

• Velocity Defense: Cross-references proposed commitment hours against your historical 7-day focus baseline.
• 1-Click Rebalanced Proposals: Automatically calculates split, trim, or reschedule options when cognitive overload is detected.
• Sovereign Override: Proceed intentionally after explicitly acknowledging overload risk.

3. ⏱️ Focus Sprint Cockpit & Sleep-Immune Telemetry

• Drift & Sleep Immunity: Uses real-world epoch timestamp delta calculation (Date.now() + remainingMs) with visibilitychange listeners. Never lags when your laptop sleeps, screen locks, or tabs background.
• Flanked Sprint Presets: 25m Sprint, 45m Deep Work, 60m Block, and 5m Recharge break options.
• Distraction Harvest Pipeline: Park fleeting thoughts during deep work and convert them with 1-click into tomorrow’s commitments.
• Harmonic Web Audio Chime: Procedurally synthesizes a warm harmonic chord (C5-E5-G5-C6) upon completion without external MP3 dependencies.

4. 🪞 Cognitive Excuse Mirror & Receipts

• Avoidance Classifier: Analyzes postponement rationalizations to detect repeating traps (Morning Illusion, Perfectionist Stalling, Underestimation).
• Historical Receipts: Surfaces past occurrences of identical justifications.
• 15-Minute Micro-Starts: Low-friction inertia-breaker sessions to overcome task initiation resistance.

5. 🤝 1-to-1 Timezone-Aware Peer Accountability

• Timezone-Aligned Midnight Resets: Partner daily feeds calibrate accurately across international boundaries.
• In-Context Discussion Threads: Discuss blockers, share proof-of-work, and unblock partners directly on individual commitment cards.
• Background Sync: TanStack Query polling for unread message indicators without UI interruption.

6. 📊 Deterministic Behavioral Analytics

• 52-Week Consistency Heatmap: Multi-level green/amber matrix focusing on long-term compounding output.
• Capacity Execution Efficiency: Tracks planned vs. actual hours delivered to compute your empirical Planning Optimism Ratio.
• Sprint Duration Sweet-Spot Curve: Highlights task completion win-rates across 25m, 45m, and 90m+ duration buckets.

7. 🤖 Day-Agnostic AI Accountability Coach & Dynamic Scheduling

• Fluid Multi-Day Temporal Horizon: Plan, query, complete, reschedule, and update commitments for any past, present, or future date (“today”, “tomorrow”, “this Friday”, “next Monday”, “3 days ago”, or ISO YYYY-MM-DD).
• Timezone Drift Immunity: Anchors all temporal reasoning to the user’s configured local timezone (ZoneId), eliminating midnight UTC boundary errors.
• Cross-Day Source & Destination Disambiguation: Intelligently separates source dates from destination dates during cross-day moves (e.g. moving Friday’s deep work sprint to next Tuesday).
• Multi-Day Cognitive Capacity Defense: Evaluates planned cognitive loads across all touched dates to proactively alert users when any single day exceeds 6 hours (> 360m).
• Undo State Fidelity: 1-tap mutation rollbacks restore exact historical dates and states, with safety boundaries preserving closed past records.

8. 📱 Mobile Ergonomics & Thumb-Zone Bottom Navigation

• Fixed Glassmorphic Bottom Navigation: Dedicated mobile navigation bar ($\le 768\text{px}$) with safe-area insets (env(safe-area-inset-bottom)).
• Glowing Quick-Add Center FAB: Instant 1-tap commitment creation reachable comfortably within the natural mobile thumb zone.
• Responsive Viewport Transitions: Top navigation for desktop screens gracefully morphs into an app-like bottom bar on handheld devices.

9. ⌨️ Power-User Keyboard Shortcuts & Zero-State Templates

• Keyboard-First Workflow: Linear/Superhuman-style single-key navigation:
  • 1, 2, 3: Instantly jump between Today, Partners, and Insights.
  • C or N: Open Create Commitment modal from anywhere.
  • ?: Toggle the Keyboard Shortcuts Cheat Sheet Modal.
  • Cmd+K / Ctrl+K: Toggle autonomous AI Coach Drawer.
  • Esc: Dismiss active modals and drawers.
• 1-Click Starter Template Chips: Empty state on the Today view features curated templates (Deep Work Sprint 90m, High-Impact Objective 60m, Partner Sync 30m, Evening Reflection 15m) to prevent blank-canvas activation friction.

10. 🍁 Kashmiri Chinar Leaf Celebration & Optimistic Undo Engine

• Chinar Leaf Particle Burst: Interactive HTML5 Canvas celebratory animation releasing golden Chinar leaves and glowing embers upon task completion (fully compliant with prefers-reduced-motion).
• Non-Blocking Optimistic Deletion: Replaced jarring browser alerts with instantaneous visual card removal and an interactive 5-second Undo SnackBar.
• Progressive Disclosure Card Architecture: Promotes the primary ⚡ Focus and 💬 Discussion actions while housing secondary operations (Edit, Postpone, Mark Missed, Delete) in a clean kebab (...) dropdown menu.

---

🏗️ Monorepo Architecture

AazDoh/
├── backend/                  # Java 21 & Spring Boot 3.3+ REST API
│   ├── src/main/java/com/aazdoh/
│   │   ├── ai/              # Spring AI clients, SSE streaming & Feasibility evaluators
│   │   ├── analytics/       # Velocity metrics, Focus Telemetry & Heatmap aggregations
│   │   ├── auth/            # JWT authentication, API key generation & User sessions
│   │   ├── commitment/      # Daily commitments & status lifecycle
│   │   ├── discussion/      # Peer comments & unread notification sync
│   │   ├── partner/         # 1-to-1 partner pairing & invitation tokens
│   │   ├── review/          # Daily reflection & excuse classifier
│   │   └── user/            # User profiles & execution personas
│   └── src/main/resources/db/migration/  # Versioned Flyway SQL migrations (V1..V19)
│
├── frontend/                 # React 18, TypeScript & Vite SPA
│   ├── src/
│   │   ├── api/             # Typed API client services
│   │   ├── components/
│   │   │   ├── agent/       # Autonomous AI Coach Drawer
│   │   │   ├── analytics/   # Heatmap, Duration Curve, Friction Matrix
│   │   │   ├── commitment/  # Commitment cards, Modals, Drag & Drop Day Phases, Stress-test UI
│   │   │   ├── common/      # Chinar leaf canvas, BrandLogo, Header, Nav, MobileBottomNav, KeyboardShortcutsModal
│   │   │   ├── focus/       # FocusSprintModal, FloatingFocusBar, Web Audio Telemetry
│   │   │   ├── partnership/ # 1:1 partner dashboard & discussion drawers
│   │   │   └── review/      # Reflection & cognitive excuse mirror modals
│   │   ├── context/         # AuthContext, ToastContext (Undo support), FocusTimerContext
│   │   ├── pages/           # LandingPage, TodayPage, PartnersPage, AnalyticsPage
│   │   └── utils/           # dateUtils, leafCelebration particle engine
│   └── public/              # Static assets & Brand favicons
│
├── cli/                      # Autonomous TypeScript Terminal CLI Agent (`aazdoh-cli`)
│   ├── src/
│   │   ├── commands/        # `today`, `done`, `run`, `chat`, `stress-test`, `focus`, `undo`, `stats`
│   │   ├── services/        # Terminal API client & SSE token streamer
│   │   └── utils/           # ASCII tables, config storage, prompt sanitization
│   └── bin/                 # Executable binaries: `aazdoh`, `az`
│
└── mcp-server/               # Official Model Context Protocol Server (`aazdoh-mcp`)
    └── src/
        ├── index.ts         # Stdio MCP Server entrypoint & Tool handlers
        └── api/             # HTTP Client connecting to AazDoh REST backend

---

💻 Terminal CLI Agent (aazdoh-cli)

The AazDoh CLI (aazdoh-cli, providing both aazdoh and az commands) brings the complete AazDoh cognitive engine directly to your command line. Engineered for terminal-first developers, it enables natural language task planning, zero-latency direct check-offs, 60-second capacity stress-testing, offline sleep-immune Pomodoro sprints, and instant 1-tap undo rollbacks.

📖 Full Guide & Recipes: See the root USAGE.md and cli/README.md.

CLI Core Highlights

• ⚡ Zero-Friction Cockpit (az): Typing az pre-renders today’s commitments table and leaves the az> prompt open for continuous commands.
• ✅ Instant Interactive Check-off (az done): Interactive checkboxes or fast keyword matching with 0 AI/LLM overhead.
• 🌅 Day Phasing Support: View and categorize tasks into 🌅 Morning, ☀️ Day, 🌙 Evening, or 📋 Anytime blocks.
• 🛡️ 60-Second Plan Feasibility Audit (az stress-test): Audits cognitive load against historical 7-day capacity before planning fallacy sets in.
• ⏱️ Offline Sleep-Immune Focus Timer (az focus): Runs in the background by default with OS desktop toast alerts, live Big ASCII digital countdown (--live), or standalone borderless popup window (--popup).
• ⏪ 1-Tap Rollbacks (az undo): Immediately revert accidental AI agent mutations or reschedules.
• 💬 Interactive AI Chief of Staff (az chat): Multi-turn conversational REPL with live token streaming and instant slash commands.

---

Installation & Setup

# Install globally from npm (exposes both 'aazdoh' and 'az' commands)
npm install -g aazdoh-cli

# Authenticate once with your API key
az login -k aazdoh_live_your_api_key_here

# (Optional) Verify version and connectivity
az --version

Credentials are saved securely to ~/.aazdoh/config.json.

---

Zero-Friction Terminal Cockpit (az)

To eliminate the friction of exiting back to bash/zsh after each command, run az with no arguments:

az

What happens:

1. Renders the ASCII Harud banner and authenticated user persona.
2. Automatically pre-fetches and renders today’s commitments table with scheduled load stats.
3. Keeps the interactive prompt az> open for continuous execution.

   ⚡ AAZDOH COGNITIVE ACCOUNTABILITY CLI
   Commit. Do. Report. Reflect. (Zero BS)

   Plan for: 2026-09-19 | User: Barkat [Persona: BALANCED]

┌──────────┬────────────────────────────────┬────────────┬──────────┬──────────────┬────────────────┬────────────────────────────┐
│ Status   │ Commitment Title               │ Duration   │ Priority │ Phase        │ Category       │ Expected Outcome           │
├──────────┼────────────────────────────────┼────────────┼──────────┼──────────────┼────────────────┼────────────────────────────┤
│ ⭕ PEND  │ PostgreSQL Connection Pooling  │ 45m        │ [HIGH]   │ 🌅 Morning   │ 🎯 DEEP FOCUS  │ HikariCP tuned & verified  │
│ ⭕ PEND  │ Spring AI Stream Integration   │ 60m        │ [MED]    │ ☀️ Day       │ 🎯 DEEP FOCUS  │ SSE stream tests passing   │
│ ✅ DONE  │ Architecture Standup & Sync    │ 30m        │ [HIGH]   │ 🌅 Morning   │ ⚡ ROUTINE     │ Roadmap aligned            │
└──────────┴────────────────────────────────┴────────────┴──────────┴──────────────┴────────────────┴────────────────────────────┘

   Summary: 1 Done | 2 Pending | 2.2h total scheduled (135m) •  OPTIMAL 

   Type instructions or commands (/today, /stress-test, /stats, /undo, /help, /exit):

az> 

---

Complete CLI Command Reference

1. aazdoh today | az list — Daily Workload & Day Phasing

Inspect today’s scheduled commitments, completion statuses, day phases, priorities, categories, and scheduled cognitive load.

# Print today's schedule table
az today

# View today's schedule and immediately open interactive check-off mode:
az today -i
# or: az today --interactive

# Inspect schedule for a specific date:
az today --date 2026-09-20

2. aazdoh done | az check | az complete — Zero-Latency Task Check-Off

Fast, interactive terminal task completion with 0 AI overhead. Directly hits the REST API for instant execution.

# Interactive multi-select mode ([Space] to toggle, [Enter] to confirm):
az done

# Quick direct keyword matching (marks matching pending task done immediately):
az done "PostgreSQL"
az check redis

3. aazdoh "" | az run "" — Natural Language Execution

Execute day-agnostic natural language instructions with automatic temporal date resolution, live SSE reasoning indicators, and action receipts.

# Mark tasks completed (today or retrospective)
az "finished PostgreSQL connection pool tuning with HikariCP benchmarked"
az "completed yesterday's DSA practice"

# Add new high-impact deep focus blocks with day phases and future dates
az "add 60m deep focus on Spring AI token streaming for afternoon"
az "add 45m of System Design on Friday morning with HIGH priority"

# Multi-day and cross-day rescheduling
az "postpone security audit to tomorrow morning"
az "move Friday's AI Reading to next Monday"

4. aazdoh chat — Interactive Multi-Turn AI Chief of Staff

Launch a persistent conversational REPL with sliding context window, live SSE token streaming, and slash commands.

az chat

5. aazdoh stress-test — 60-Second Plan Feasibility Audit

Audits your daily workload against your historical 7-day velocity baseline to flag cognitive overload, task fragmentation, and planning fallacy.

# Run standard audit
az stress-test

# Pass a defense justification to re-evaluate:
az stress-test --defense "No meetings today, dedicated uninterrupted focus block"

6. aazdoh focus | az timer — Offline Pomodoro, Big ASCII Clock & Popout Window

Run a 100% offline, zero-network focus timer. Runs in the background by default, freeing your terminal immediately, and triggers native OS desktop toast notifications upon completion.

# 25-minute Pomodoro in background (frees terminal immediately)
az focus

# Custom duration & task in background
az focus 45m "PostgreSQL pool optimization"
az focus 1.5h "Drafting system architecture"

# Check active timer status & ASCII progress bar
az focus status

# Stop / cancel active background timer
az focus stop

# Interactive live full-screen countdown with Big ASCII Digital Clock & hotkeys:
# ([Space] Pause/Resume, [+] +5m, [-] -5m, [w/p] Popout Window, [q] Exit)
az focus 25m --live

# Launch a sleek floating desktop popup clock window (borderless app mode)
az focus 45m "Core Engine Architecture" --popup

7. aazdoh undo — Instant Mutation Rollback

Instantly rollback the most recent action executed by the AI Agent (task creation, status change, or rescheduling).

az undo

8. aazdoh stats | az velocity — 7-Day Telemetry & Velocity

Displays 7-day compounding consistency, total focus hours, planning optimism ratio, and completion breakdown.

az stats
az stats --days 14

---

Interactive REPL Slash Commands

Inside the interactive cockpit (az>) or az chat, use the following slash commands:

---

Configuration & Credential Priority

The CLI resolves configuration through a strict 5-layer hierarchy (highest to lowest):

1. CLI Runtime Flags       (--key , --url )
2. System Environment      (export AAZDOH_API_KEY=aazdoh_live_...)
3. Directory .env          (AAZDOH_API_KEY in current working directory)
4. Local User Config       (~/.aazdoh/config.json via `az login`)
5. Hardcoded Default       (https://aazdoh.onrender.com)

Configuration File (~/.aazdoh/config.json):

{
  "apiUrl": "https://aazdoh.onrender.com",
  "apiKey": "aazdoh_live_...",
  "userFullName": "Barkat",
  "userEmail": "[email protected]",
  "aiPersona": "BALANCED"
}

---

Security & Input Guardrails

• 500-Character Universal Input Ceiling: All prompt inputs and natural language instructions are capped at 500 characters to prevent prompt injection and oversized payload attacks.
• Delimiter Neutralization: Backticks and markdown block delimiters are sanitized before passing to backend LLMs.
• Zero Local Credential Leakage: Configuration is stored in OS-isolated user directories (~/.aazdoh) with appropriate file permissions.
• Reversible Mutations: All agent executions are recorded with reversible log receipts for 1-click rollback via az undo.

---

🤖 Model Context Protocol (aazdoh-mcp)

The AazDoh MCP Server (aazdoh-mcp) is an official Model Context Protocol server that directly bridges your AI coding assistants with your personal accountability system. It turns coding assistants in Cursor, Claude Desktop, and Antigravity into proactive AI Chiefs of Staff capable of auditing cognitive load, locking in focus blocks before refactoring, validating proof-of-work, and updating accountability partners.

📖 Full Guide & Tool Schemas: See mcp-server/README.md and USAGE.md.

                      ┌─────────────────────────────────────────┐
                      │          AazDoh MCP Server              │
                      └────────────────────┬────────────────────┘
                                           │
         ┌──────────────────┬──────────────┴─────┬──────────────────┐
         │                  │                    │                  │
┌────────▼────────┐ ┌───────▼────────┐ ┌─────────▼────────┐ ┌───────▼────────┐
│ Plan & Commit   │ │ AI Intelligence│ │ Daily Reflection │ │ Partner Feeds  │
│ 8 Tools         │ │ 6 Tools        │ │ 2 Tools          │ │ 4 Tools        │
└─────────────────┘ └────────────────┘ └──────────────────┘ └────────────────┘

---

Supported AI Environments

• Antigravity IDE
• Cursor IDE
• Claude Desktop
• Windsurf / Cascade
• VS Code (via Cline, Roo Code, Claude Dev)

---

IDE Configuration Snippets

1. Antigravity IDE & Cursor (mcp_config.json or .cursor/mcp.json)

{
  "mcpServers": {
    "aazdoh": {
      "command": "npx",
      "args": ["-y", "aazdoh-mcp@latest"],
      "env": {
        "AAZDOH_API_KEY": "aazdoh_live_your_api_key_here",
        "AAZDOH_API_URL": "https://aazdoh.onrender.com"
      }
    }
  }
}

2. Claude Desktop (claude_desktop_config.json)

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "aazdoh": {
      "command": "npx",
      "args": ["-y", "aazdoh-mcp@latest"],
      "env": {
        "AAZDOH_API_KEY": "aazdoh_live_your_api_key_here",
        "AAZDOH_API_URL": "https://aazdoh.onrender.com"
      }
    }
  }
}

---

Complete Suite of 18 Native MCP Tools

The MCP server exposes 18 specialized tools organized into 4 functional domains:

🎯 1. Plan & Commitment Execution (8 Tools)

🛡️ 2. AI Behavioral Intelligence & Stress-Testing (6 Tools)

🪞 3. Daily Retrospective & Reflection (2 Tools)

🤝 4. Peer Accountability & Discussion Threads (4 Tools)

---

AI Prompting Recipes & Workflows

Once aazdoh-mcp is configured in your IDE, use these practical prompt patterns during development:

🌅 Recipe 1: Morning Standup & Cognitive Load Audit

“Check my AazDoh today’s plan using get_today_plan. If my scheduled cognitive load exceeds 4.5 hours, run stress_test_plan and help me prioritize the 2 highest-leverage deep focus tasks for our coding session.”

⏱️ Recipe 2: Deep Focus Sprint Lock-In Before Refactoring

“We’re about to refactor the database connection pool. Lock in a 45m DEEP_WORK commitment in AazDoh titled ‘PostgreSQL Pool Tuning’ with HIGH priority in the MORNING phase and an expected outcome of ‘Passing concurrency stress tests’.”

✅ Recipe 3: Task Completion & Partner Proof-of-Work

“We just fixed the token refresh race condition and all unit tests passed. Mark the authentication commitment complete in AazDoh and post a partner update to the thread with a summary of what we changed.”

🪞 Recipe 4: Evening Retrospective & Friction Analysis

“Review today’s completed vs missed commitments in AazDoh. Summarize our velocity and submit our daily reflection with insights on what caused friction during the afternoon block.”

---

Daily Developer Operational Playbook

---

🚀 Quickstart & Installation

Prerequisites

• Java JDK 21+
• Node.js 18+ & npm
• PostgreSQL 15+
• (Optional) OpenAI / LLM API Key for AI Feasibility & Behavioral Synthesis

---

1. Database Setup

Create a local PostgreSQL database:

CREATE DATABASE aazdoh;

---

2. Backend Configuration

1. Navigate to backend/:

   cd backend
   

2. Set environment variables or configure src/main/resources/application.yml:

   export SPRING_DATASOURCE_URL=jdbc:postgresql://localhost:5432/aazdoh
   export SPRING_DATASOURCE_USERNAME=postgres
   export SPRING_DATASOURCE_PASSWORD=postgres
   export JWT_SECRET=your_super_secret_jwt_key_at_least_256_bits_long
   export SPRING_AI_OPENAI_API_KEY=your_openai_api_key # Optional
   

3. Run Spring Boot:

   # Linux / macOS
   ./mvnw spring-boot:run
   
   # Windows PowerShell
   .\mvnw.cmd spring-boot:run
   

   The backend starts on port 8080. Flyway automatically executes database migrations.

---

3. Frontend Setup

1. Navigate to frontend/:

   cd frontend
   

2. Install dependencies and start the Vite dev server:

   npm install
   npm run dev
   

3. Open your browser at http://localhost:3000.

---

4. CLI Setup

cd cli
npm install
npm run build
npm link # Symlink globally for local testing

---

5. MCP Server Setup

cd mcp-server
npm install
npm run build

---

🧪 Testing & Build Verification

# Backend unit & integration tests
cd backend
./mvnw test

# Frontend TypeScript check & production bundle
cd frontend
npm run build

# CLI compilation
cd cli
npm run build

# MCP server compilation
cd mcp-server
npm run build

---

🎨 Design System (Kashmir Harud Aesthetic)

AazDoh features a bespoke, artisanal design system inspired by the autumn season (Harud) of the Kashmir valley, tuned for low eye-strain deep work and tactile responsiveness:

Micro-Interactions & Usability Ergonomics

• 🍁 Canvas Leaf Particle Engine: Real-time celebratory particle simulation rendering spinning Chinar leaves and golden embers upon task completion.
• 📱 Mobile Thumb Zone: Fixed bottom navigation on viewports $\le 768\text{px}$ with center quick-add FAB and safe-area inset padding.
• ⌨️ Vim/Linear-Style Navigation: 1-3 route jumping, C/N quick add, ? shortcuts sheet, Cmd+K AI coach.
• ⚡ Optimistic Deletion with Undo: Immediate UI responsiveness backed by a 5-second reversible SnackBar window.

---

📄 License

This project is licensed under the MIT License — see the LICENSE file for details.