FocusRoom | Productivity Agent

A personal productivity workspace built with Python, Streamlit, Groq, SQLite, and MCP. It combines task management, daily planning, persistent memories, a multi-agent supervisor, and optional email reminders in one project.

Features

• Streamlit productivity dashboard
• Create, list, filter, and complete tasks
• Task priorities, due dates, projects, and statuses
• Persistent SQLite storage in productivity.db
• Daily plans and productivity reports
• Save and search personal memories
• Groq-powered productivity assistant
• Multi-agent supervisor for task, memory, and planning requests
• MCP server with HTTP tools for external clients
• Optional automatic email reminders

Requirements

• Python 3.10 or newer
• A Groq API key for the AI features
• Gmail SMTP credentials or another SMTP provider for email reminders

Setup

1. Create a virtual environment

PowerShell:

python -m venv .venv
.\.venv\Scripts\Activate.ps1

If PowerShell blocks script execution for the current session:

Set-ExecutionPolicy -Scope Process -ExecutionPolicy RemoteSigned
.\.venv\Scripts\Activate.ps1

2. Install dependencies

python -m pip install -r requirements.txt

3. Configure environment variables

Copy .env.example to .env and fill in the values:

Copy-Item .env.example .env

Required for the Groq-backed agent:

GROQ_API_KEY=your_groq_api_key
GROQ_MODEL=openai/gpt-oss-120b

Optional MCP configuration:

MCP_HOST=127.0.0.1
MCP_PORT=8000
MCP_API_KEY=local_or_remote_api_key

For email reminders:

EMAIL_SENDER=your_email@gmail.com
EMAIL_RECIPIENT=recipient@example.com
EMAIL_PASSWORD=your_smtp_or_app_password
SMTP_SERVER=smtp.gmail.com
SMTP_PORT=587

Do not commit .env or API credentials. They are excluded by .gitignore.

Run the Streamlit app

streamlit run streamlit_app.py

The app provides these sections:

• Overview - task metrics, active work, daily snapshot, and completion progress
• Tasks - create and filter tasks, then mark tasks complete
• Plan - view the current daily plan and productivity report
• Memory - save and search persistent context
• Assistant - send natural-language requests to the productivity supervisor

Run the MCP server

Start the Streamable HTTP MCP server with:

python -m orchestrator.mcp_server

By default, it runs on http://127.0.0.1:8000/mcp.

The server exposes tools for:

• productivity_assistant
• create_task
• list_tasks
• update_task
• complete_task
• delete_task
• save_memory
• search_memory
• daily_plan
• productivity_report

When binding the MCP server to a non-local host, set MCP_API_KEY. Remote requests must use a bearer token:

Authorization: Bearer <MCP_API_KEY>

Run reminders

The reminder worker checks upcoming tasks and sends configured email notifications:

python reminder_worker.py

The reminder schedule is configured in the worker and supports reminders at 24 hours, 1 hour, and 15 minutes before a task is due. Email reminders require valid SMTP settings in .env.

Run tests

Use the project virtual environment so pytest is available:

.\.venv\Scripts\python -m pytest -q

To run a specific test file:

.\.venv\Scripts\python -m pytest -q test_multi_agent.py

To check Python syntax without starting the app:

.\.venv\Scripts\python -m compileall streamlit_app.py orchestrator

Project structure

Productivity_Agent/
|-- streamlit_app.py              # Streamlit user interface
|-- requirements.txt              # Python dependencies
|-- .env.example                  # Environment variable template
|-- productivity.db               # Local SQLite database, generated locally
|-- reminder_service.py           # Reminder service implementation
|-- reminder_worker.py            # Continuous reminder worker
|-- orchestrator/
|   |-- agent.py                  # Core task, memory, plan, and report logic
|   |-- database.py               # SQLite persistence layer
|   |-- orchestrator.py           # Multi-agent supervisor
|   |-- mcp_server.py             # MCP HTTP server and tools
|   |-- memory_agent.py           # Memory agent adapter
|   |-- planning_agent.py         # Planning agent adapter
|   |-- task_agent.py             # Task agent adapter
|-- test_*.py                     # Project tests

Data and security notes

• Tasks and memories are stored locally in SQLite.
• .env contains secrets and must remain private.
• For Gmail, use an app password where required instead of your primary account password.
• The MCP API key is required when the server is exposed beyond localhost.
• Back up productivity.db if the local task and memory history is important.

Troubleshooting

pytest is not recognized

Run pytest through the virtual environment:

.\.venv\Scripts\python -m pytest -q

The app starts but the assistant is unavailable

Check that .env exists and contains a valid GROQ_API_KEY, then restart Streamlit.

Streamlit reports missing ScriptRunContext

This warning appears when importing a Streamlit module directly with python -c. Start the application with streamlit run streamlit_app.py for normal operation.

Compilation works but tests fail during collection

A collection error means pytest could not finish importing the tests. Read the first reported exception and fix that dependency or constructor mismatch before evaluating the remaining tests.