DevFlow AI is a productivity assistant designed to help software developers manage tasks, schedule work sessions, and optimize their workflow through natural language interaction and integrated tools. It combines real-world APIs (Google Calendar, Google Tasks, Github) with an agentic reasoning engine to understand developer intent and automate planning, reflection, and execution workflows.

Documentation

See below for quickstart installation and usage examples. For comprehensive guidance on setting up and using DevFlow AI,

Prerequisites

• Python 3.10+ installed.
• Ollama installed and running (for the local LLM).
• Google Cloud Credentials (JSON key file)
• Google APIs Enabled: Ensure that Google Calendar and Google Tasks APIs are enabled in your Google Cloud project.
• GitHub Personal Access Token: Create a PAT with repo and workflow scopes.

Setup the AI (Ollama)

We use llama3.1:8b for its superior tool-calling capabilities.

1. Open your terminal and start Ollama (or open the desktop app):

   ollama serve

2. In a new terminal tab, pull the specific model:

   ollama pull llama3.1:8b

---

Clone & Install

# Clone the repository
git clone https://github.com/being-bilal/devflow-ai.git

cd devflow-ai

# Create a virtual environment 
python -m venv venv

# Activate it
source venv/bin/activate

# Install dependencies
pip install -r requirements.txt

---

Configure Environment & Secrets

DevFlow AI requires API keys to function.

1. Create .env File Copy the example template and add your keys (GitHub Token, etc.):

# Copy the example environment file
cp .env.example .env

2. Google Auth Setup Place your Google Cloud JSON key file (e.g., service_account.json) in the root folder of the project.
3. Update Config Open your .env file and ensure the GOOGLE_APPLICATION_CREDENTIALS variable matches the exact filename of your JSON key.
4. GitHub Token Generate a GitHub Personal Access Token (PAT) with repo and workflow scopes and add it to the GITHUB_TOKEN variable in your .env file.
5. Langfuse Setup (Optional) Sign up at Langfuse and obtain your public and secret keys. Add them to the respective variables in your .env file to enable observability

Run DevFlow AI

You need to run the Backend and Frontend in two separate terminals.

1. Step A: Start the Backend API (Terminal 1) : Open your terminal, activate your environment, and run this command in your project root:

python api_server.py

• Keep this terminal open.

2. Start the Web UI (Terminal 2): Open a new terminal window, activate your environment, and run:

cd web
python -m http.server 3000

• Open your browser and navigate to http://localhost:3000 to access the DevFlow AI web interface.

Architecture Overview

The DevFlow AI system is designed as a modular and scalable developer productivity agent. The components interact as follows:

• User: Sends natural language requests to the system through the FastAPI REST API.
• FastAPI REST API: Serves as the entry point, handling incoming requests, authentication, and authorization.
• LangChain Agent: Acts as the reasoning engine, interpreting user intent and determining which tools to invoke.
• MCP Tool Orchestrator: Executes the agent’s decisions, coordinating interactions with external services and APIs.
• APIs Integrated:
  • Google Calendar API — schedules events and manages calendar entries.
  • Google Tasks API — creates, updates, and tracks tasks.
  • GitHub API — retrieves repository data and manages code-related workflows.
• Langfuse Observability: Collects logs, metrics, and traces of agent operations to monitor performance and improve reliability.

The flow ensures that user commands are processed intelligently, executed through the appropriate tools, and monitored for observability, providing a seamless productivity experience for developers.

Demo

The following screenshots demonstrate key capabilities of DevFlow AI, from natural language interaction to real-world API execution and observability.

1. Natural Language Request

2. Model & Settings Configuration

3. Google Tasks and Calendar Integration

4. GitHub Integration

5. Observability & Tracing with Langfuse

Future Improvements

DevFlow AI is designed with extensibility in mind. Potential areas for future enhancement include:

• Expanded Tooling Ecosystem
  Support for additional developer tools such as Jira, Linear, Slack, and CI/CD platforms to further streamline workflows.

• Long-Term Memory & Personalization
  Introducing persistent memory to allow the agent to learn user preferences, work patterns, and long-term goals over time.

• Advanced Planning & Prioritization
  Smarter task prioritization using historical data, deadlines, and workload analysis.

• Security & Permissions Control
  Security access control for connected services to support team and enterprise use cases.