Logfire AI Assistant by Vstorm

AI-Powered Debugging and Analysis for Pydantic Logfire
_{Built by Vstorm — AI tooling for modern development teams}

Quick Start • Features • Screenshots • How It Works

🔍 Natural Language Queries  •  🗃️ Auto SQL Generation  •  📊 Smart Visualizations  •  ⚡ Real-time Streaming

---

Logfire AI Assistant by Vstorm is an AI-powered tool that helps you debug, analyze, and understand your applications using Pydantic Logfire. Ask questions in plain English, get instant answers with SQL queries and visualizations. Developed by Vstorm as part of our open-source AI developer tools ecosystem.

Ask questions in natural language, get SQL queries and visualizations

Why Use It?

Debugging AI agents and complex applications is hard:

• Long traces — Agents generate many spans across multiple tool calls and LLM requests
• Complex prompts — System prompts can be lengthy, making it hard to spot issues
• Multi-turn conversations — Understanding what happened requires looking across entire interaction histories

Vstorm's Logfire AI Assistant acts as your AI debugging partner — it understands agent architectures, recognizes failure patterns, and helps you ship better software faster.

---

📸 Screenshots

Main Interface

Chrome extension sidebar integrated with Logfire dashboard

Core Features

Query Results	SQL Generation
AI responses with data tables	View generated SQL queries

Chart Visualization	Span Analysis
Auto-generated charts from query results	Select spans directly in Logfire UI

Settings & Configuration

Multi-Project Support	Custom Prompts
Switch between Logfire projects	Create reusable prompt templates

LLM Model Selection	Customization
Choose from multiple AI models	Customize appearance and behavior

Additional Features

Slash Commands	Conversation History
Quick access to saved prompts	Browse and continue past conversations

File Attachments
Attach files for context

---

✨ Features

Feature	Description
🔍 Natural Language Queries	Ask about your app's behavior in plain English
🗃️ SQL Query Generation	AI translates questions into SQL against Logfire data
📊 Auto Visualizations	Automatically generate charts from query results
🎯 Span Context	Select spans in Logfire and ask "what happened here?"
📁 Multi-Project Support	Switch between different Logfire projects seamlessly
📝 Custom Prompts	Define reusable prompts for common analysis tasks
⚡ Real-time Streaming	Responses stream via WebSocket as they're generated
🔐 JWT Authentication	Secure access with token-based auth

---

🚀 Quick Start

1. Backend

# Install & configure
make install
cp backend/.env.example backend/.env
# Edit backend/.env with your settings

# Start database & run migrations
make db-init && make db-upgrade

# Run server
make run dev

2. Browser Extension

cd browser-extension
bun install && bun run build

Then load in Chrome:

1. Go to chrome://extensions/
2. Enable Developer mode
3. Click Load unpacked → select browser-extension/dist

That's it! Open Logfire dashboard, click on any span, and start asking questions.

---

🔄 How It Works

You: "Why did this request take so long?"
                    │
                    ▼
┌─────────────────────────────────────────────────────────┐
│  Chrome Extension                                       │
│  • Captures span context (trace_id, span_id)            │
│  • Sends question via WebSocket                         │
└─────────────────────────────────────────────────────────┘
                    │
                    ▼
┌─────────────────────────────────────────────────────────┐
│  AI Agent (PydanticAI)                                  │
│  • Analyzes question + context                          │
│  • Generates SQL query                                  │
│  • Executes against Logfire API                         │
│  • Creates visualization if needed                      │
└─────────────────────────────────────────────────────────┘
                    │
                    ▼
Answer: "The database query on line 42 took 3.2s due to
        missing index on user_id column. Here's the
        breakdown chart..."

---

🏗️ Architecture

┌─────────────────────────────────────────────────────────────┐
│                    Chrome Extension                         │
│  ┌─────────────────┐  ┌──────────────────────────────────┐  │
│  │ Content Script  │  │         Sidebar Panel            │  │
│  │ (Logfire page)  │  │  • Chat interface                │  │
│  │ • Span selector │  │  • Project switcher              │  │
│  │ • Context grab  │  │  • Settings & prompts            │  │
│  └────────┬────────┘  └──────────────┬───────────────────┘  │
└───────────┼──────────────────────────┼──────────────────────┘
            │                          │
            │      WebSocket           │
            └──────────┬───────────────┘
                       ▼
┌─────────────────────────────────────────────────────────────┐
│                      Backend (FastAPI)                      │
│  ┌──────────────┐  ┌──────────────┐  ┌──────────────────┐   │
│  │   API Routes │  │   Services   │  │   AI Agent       │   │
│  │   /api/v1/*  │──│              │──│   (PydanticAI)   │   │
│  └──────────────┘  └──────────────┘  └────────┬─────────┘   │
│                                               │             │
│  ┌────────────────┐  ┌──────────────┐  ┌──────▼───────────┐ │
│  │  PostgreSQL    │  │    Redis     │  │  Logfire API     │ │
│  │  (users,       │  │  (cache,     │  │  (query data)    │ │
│  │  conversations)│  │  rate limit) │  └──────────────────┘ │
│  └────────────────┘  └──────────────┘                       │
└─────────────────────────────────────────────────────────────┘

---

🛠️ Development

Backend Commands

make run dev           # Development server
make test              # Run tests
make test-cov          # Tests with coverage
make lint              # Check linting
make format            # Auto-fix formatting
make db-migrate        # Create new migration
make docker-up         # Start all services (Docker)

Extension Commands

cd browser-extension
bun run dev            # Watch mode
bun run build          # Production build
bun run check          # Lint + format (Biome)

---

📁 Project Structure

├── backend/
│   └── app/
│       ├── agents/           # AI agent (PydanticAI)
│       ├── api/routes/v1/    # HTTP endpoints
│       ├── services/         # Business logic
│       ├── repositories/     # Data access
│       ├── schemas/          # Pydantic models
│       └── db/models/        # SQLAlchemy models
├── browser-extension/
│   └── src/
│       ├── background/       # Service worker
│       ├── content/          # Content script
│       ├── sidebar/          # Side panel UI
│       └── stores/           # Zustand stores
└── docker-compose.yml

---

⚙️ Environment Variables

Copy backend/.env.example to backend/.env:

ENVIRONMENT=local
SECRET_KEY=your-secret-key     # openssl rand -hex 32

POSTGRES_HOST=localhost
POSTGRES_PASSWORD=postgres

REDIS_HOST=localhost

LOGFIRE_TOKEN=your-token       # Optional, for backend observability

---

🙏 Built With

• Pydantic Logfire — Observability platform from the Pydantic team
• Pydantic AI — The agent framework powering this assistant
• Inspired by Polly by LangChain — AI Agent Engineer for LangSmith

---

📄 License

See LICENSE for details.

---

Made with ❤️ by Vstorm