Vstorm-co has introduced a full-stack AI agent template designed to streamline the creation of AI-powered applications. The template leverages FastAPI for the backend and Next.js for the frontend, offering a robust architecture for AI agents, Retrieval Augmented Generation (RAG), real-time streaming, and user authentication. It comes with over 20 integrations, making it suitable for various enterprise use cases.
The template supports six different AI agent frameworks: PydanticAI, PydanticDeep, LangChain, LangGraph, CrewAI, and DeepAgents. It also includes a RAG pipeline with support for vector databases like Milvus, Qdrant, pgvector, and ChromaDB. Key features include WebSocket streaming for real-time chat UIs, conversation sharing capabilities, and enterprise-grade readiness with JWT, OAuth, an admin panel, Celery, Docker, and Kubernetes support.
The latest release, version 0.2.9, introduces a chart generation tool that allows agents to produce various chart types as JSON, which are then rendered interactively in the web chat or as PNGs for other platforms. It also adds a portable, SSRF-safe `fetch_url` tool for web page retrieval across multiple agent frameworks and refreshes default AI models to include newer versions like OpenAI's `gpt-5.5` and Anthropic's `claude-opus-4-7`. The agent system prompt has been rewritten for an outcome-first style, and the `web_search` tool now returns structured JSON for improved UI rendering.