Skip to content

amyemk/coco

Repository files navigation

AI Chief of Staff

A personal AI assistant designed for Chief Product Officers that delivers intelligent daily briefings, helps prioritize work, drafts email replies, and maintains continuous understanding of your product organization.

Overview

The AI Chief of Staff sits across strategy, execution, and communicationβ€”translating vision into action while reducing cognitive load and context switching. It synthesizes information from Gmail, Google Calendar, and Coda to provide:

  • Daily Briefings: Smart morning digests with priorities, email drafts, and key updates
  • Email Assistance: AI-generated reply drafts that match your tone and context
  • Meeting Intelligence: Automated follow-ups and action item extraction
  • Strategic Context: Continuous tracking of decisions, roadmaps, and OKRs

Key Features

Daily Briefing (7 AM delivery)

  • πŸ—‚οΈ Today's Priorities: Synthesized from calendar, emails, and Coda docs
  • πŸ“¬ Smart Email Drafts: Suggested replies for high-priority messages
  • πŸŽ™οΈ Meeting Follow-ups: Action items from recent meetings
  • πŸ“Š Product Org Updates: Key changes to roadmaps, OKRs, strategy docs

Intelligent Assistance

  • Priority Scoring: AI-powered email and task prioritization
  • Context Awareness: Semantic understanding of your org's decisions and "why"
  • Change Detection: Alerts when important documents are modified
  • Inconsistency Flagging: Surfaces tensions between decisions and commitments

Architecture

AI Chief of Staff
β”œβ”€β”€ Daily Briefing Generator
β”œβ”€β”€ Context Engine (maintains org knowledge)
β”œβ”€β”€ Action Engine (prioritizes and generates drafts)
└── AI/LLM Layer (powered by Claude)

Integrations:
β”œβ”€β”€ Gmail (OAuth 2.0)
β”œβ”€β”€ Google Calendar (OAuth 2.0)
└── Coda (API token)

Storage:
β”œβ”€β”€ SQLite (structured data)
└── ChromaDB (vector embeddings)

Technology Stack

  • Language: Python 3.11+
  • AI: Anthropic Claude API
  • Database: SQLite with FTS5
  • Vector Store: ChromaDB
  • Scheduler: APScheduler
  • APIs: Gmail, Google Calendar, Coda

Quick Start

Prerequisites

  • Python 3.11 or higher
  • Google Cloud account (for Gmail/Calendar APIs)
  • Anthropic API key
  • Coda account (optional)

Installation

# Clone the repository
git clone <repository-url>
cd coco

# Create virtual environment
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# Set up configuration
cp .env.example .env
cp config/config.example.yaml config/config.yaml

# Edit .env and add your API keys
# Edit config.yaml with your preferences

# Initialize database
python scripts/init_db.py

# Authenticate with Google
python scripts/authenticate.py

Configuration

  1. Google Cloud Setup:

  2. Anthropic API:

  3. Coda (optional):

    • Generate API token at https://coda.io/account
    • Add to .env as CODA_API_TOKEN
    • Configure tracked docs in config.yaml

Usage

# Run manual sync
python scripts/sync_all.py

# Generate briefing manually (for testing)
python scripts/trigger_briefing.py

# Start the scheduler (runs daily at 7 AM)
python src/main.py

Project Structure

coco/
β”œβ”€β”€ src/                      # Source code
β”‚   β”œβ”€β”€ ai/                   # AI/LLM integration
β”‚   β”œβ”€β”€ auth/                 # Authentication
β”‚   β”œβ”€β”€ briefing/             # Daily briefing generation
β”‚   β”œβ”€β”€ config/               # Configuration management
β”‚   β”œβ”€β”€ db/                   # Database setup and migrations
β”‚   β”œβ”€β”€ delivery/             # Email/Slack delivery
β”‚   β”œβ”€β”€ engine/               # Core logic (scoring, generation)
β”‚   β”œβ”€β”€ integrations/         # External API integrations
β”‚   β”‚   β”œβ”€β”€ gmail/
β”‚   β”‚   β”œβ”€β”€ calendar/
β”‚   β”‚   └── coda/
β”‚   β”œβ”€β”€ models/               # Data models
β”‚   β”œβ”€β”€ repositories/         # Data access layer
β”‚   β”œβ”€β”€ scheduler/            # Job scheduling
β”‚   β”œβ”€β”€ sync/                 # Data synchronization
β”‚   └── utils/                # Utilities
β”œβ”€β”€ tests/                    # Test suite
β”‚   β”œβ”€β”€ unit/
β”‚   β”œβ”€β”€ integration/
β”‚   β”œβ”€β”€ e2e/
β”‚   └── fixtures/
β”œβ”€β”€ scripts/                  # Utility scripts
β”œβ”€β”€ docs/                     # Documentation
β”œβ”€β”€ data/                     # Local data storage
β”œβ”€β”€ logs/                     # Application logs
β”œβ”€β”€ config/                   # Configuration files
β”œβ”€β”€ requirements.txt          # Python dependencies
β”œβ”€β”€ PRD.md                    # Product requirements
β”œβ”€β”€ TECHNICAL_SPEC.md         # Technical specification
β”œβ”€β”€ ARCHITECTURE.md           # System architecture
β”œβ”€β”€ IMPLEMENTATION_ROADMAP.md # Development roadmap
└── README.md                 # This file

Documentation

Specifications

Setup Guides

Future Documentation

  • User Guide (Phase 3)
  • Configuration Reference (Phase 3)
  • API Documentation (Phase 5)

Development

Running Tests

# Run all tests
pytest

# Run specific test suite
pytest tests/unit/
pytest tests/integration/

# Run with coverage
pytest --cov=src --cov-report=html

Code Quality

# Format code
black src/ tests/

# Lint code
ruff check src/ tests/

# Type checking
mypy src/

Roadmap

βœ… Phase 1: Foundation (Weeks 1-2) - COMPLETE

  • βœ… Project setup and configuration
  • βœ… OAuth authentication with Google
  • βœ… Database initialization with SQLite + FTS5
  • βœ… Configuration management system
  • βœ… Setup scripts and tests

🚧 Phase 2: Data Integration (Weeks 2-3) - NEXT

  • Gmail sync service
  • Calendar sync service
  • Coda sync service
  • Incremental update logic

⏳ Phase 3: AI Integration (Weeks 3-4)

  • Claude API client
  • Priority scoring
  • Email draft generation
  • Task extraction

⏳ Phase 4: Daily Briefing MVP (Weeks 4-5)

  • Briefing generation
  • Email delivery
  • Scheduled jobs

⏳ Phase 5: Polish & Testing (Weeks 5-6)

  • Real data testing
  • UX improvements
  • Performance optimization

⏳ Phase 6: Iteration (Weeks 6+)

  • Feedback-based improvements
  • Advanced features
  • Optional web UI

See IMPLEMENTATION_ROADMAP.md for detailed timeline.

Security & Privacy

  • Local Storage: All data stored locally in encrypted SQLite database
  • No Third-Party Sharing: Data only sent to APIs you explicitly configure
  • Secure Credentials: OAuth tokens and API keys encrypted at rest
  • Data Retention: Configurable retention policies for emails, events, docs
  • Privacy First: No telemetry, no usage tracking, full user control

Success Metrics

  • 80% briefing open rate within first hour

  • 50% of AI-generated replies used or edited

  • At least 3 actionable tasks surfaced per day
  • Subjective: "Saves me >1 hour per day"

Contributing

This is currently a personal project. If you'd like to contribute or adapt for your own use, please:

  1. Fork the repository
  2. Create a feature branch
  3. Make your changes
  4. Submit a pull request

License

[To be determined]

Support

For issues, questions, or feedback:

  • Open an issue on GitHub
  • Email: [your-email]

Acknowledgments

  • Built with Anthropic Claude
  • Inspired by the need for better executive productivity tools
  • Thanks to the open-source community

Status: Phase 1 Complete βœ… | Phase 2 Ready πŸš€

Last Updated: 2026-01-09

Phase 1 Achievements:

  • Complete database schema with 10+ tables and FTS5 search
  • Google OAuth authentication (Gmail + Calendar)
  • Configuration management with YAML + env vars
  • Automated setup scripts (init_db.py, authenticate.py, check_status.py)
  • Comprehensive test suite with 15+ unit tests
  • Full documentation and troubleshooting guides

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages