jupyter-ai-persona-manager

What it is and what it does

jupyter_ai_persona_manager is the core registry and lifecycle manager for AI personas in Jupyter AI chat environments. It provides BasePersona (an abstract base class for custom assistants), PersonaManager (for registration and lifecycle), and entry-point-based discovery so that personas can be packaged as plugins and automatically discovered. Each persona is analogous to a bot in other chat systems—a distinct AI assistant with its own behavior, model, and capabilities that can coexist in the same chat.

Developers use it to create custom personas by subclassing BasePersona, implementing a process_message method, and registering via Python entry points or by placing persona files in .jupyter/personas/. The package handles avatar serving, multi-user awareness, and automatic discovery. It integrates with jupyterlab-chat and jupyter-server, making it the foundation for extending Jupyter AI with custom assistants.

Use it for:

• Create a custom code-review assistant persona that analyzes notebook cells and provides feedback on code quality and style.
• Build a research-focused persona that integrates with external APIs (e.g., Ollama, OpenAI) to provide domain-specific answers in Jupyter chat.
• Develop a local development helper persona loaded from .jupyter/personas/ for iterative testing without restarting JupyterLab.
• Register multiple personas via entry points so users can @-mention different AI assistants for different tasks in the same chat.
• Extend Jupyter AI with a persona that queries internal documentation or knowledge bases specific to an organization.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Provides the foundational infrastructure for managing and registering AI personas (custom chat assistants) in Jupyter AI environments, including base classes, lifecycle management, and automatic discovery via entry points.

Yes, if you are running JupyterLab 4.0+ and want to create or use custom AI personas in Jupyter AI chat. The package is actively maintained, has low install friction, carries no security vulnerabilities, and is permissively licensed. Install it as a dependency when building persona plugins, or as a base package if you intend to develop custom personas locally.

Install

jupyter-ai-persona-manager on PyPI

pip install jupyter-ai-persona-manager

uv add jupyter-ai-persona-manager

poetry add jupyter-ai-persona-manager

Installing jupyter-ai-persona-manager

Before you install

Low install friction with a pure-Python wheel. Active maintenance (last commit 2026-08-12, released 2026-07-23) and current Python support (3.10–3.14). Depends on jupyter-server, jupyterlab-chat, and other Jupyter ecosystem packages, all standard in JupyterLab environments.

License in practice

BSD 3-Clause License permits use, modification, and distribution with minimal restrictions—suitable for both open-source and proprietary projects. No copyleft obligations.

Quickstart

pip install jupyter_ai_persona_manager

from jupyter_ai_persona_manager import BasePersona, PersonaDefaults
from jupyterlab_chat.models import Message

class MyPersona(BasePersona):
    @property
    def defaults(self):
        return PersonaDefaults(
            name="MyPersona",
            description="Custom assistant",
            system_prompt="You are helpful."
        )
    
    async def process_message(self, message: Message):
        self.send_message(f"You said: {message.body}")

Requires JupyterLab >= 4.0.0 and Python >= 3.10. Personas are discovered via entry points or from .jupyter/personas/ directory; restart JupyterLab or use /refresh-personas command to load new personas.

Verify before relying

• Whether avatar serving at /api/ai/avatars/{filename} works reliably when multiple personas share the same filename.
• Performance characteristics when many personas are registered or when processing high-volume messages.
• Compatibility with Jupyter AI versions other than the latest.

Package facts

Evidence: jupyter_ai_persona_manager-0.1.2-py3-none-any.whl

Keywords: jupyter, jupyterlab, jupyterlab-extension

Tags

More Application Frameworks packages

Further reading