jupyter_ai_persona_manager

The core manager & registry for AI personas in Jupyter AI.

This package provides the foundational infrastructure for managing AI personas in Jupyter AI chat environments. It includes:

• BasePersona: Abstract base class for creating custom AI personas
• PersonaManager: Registry and lifecycle management for personas
• PersonaAwareness: Awareness integration for multi-user chat environments
• Entry Point Support: Automatic discovery of personas via Python entry points

AI personas are analogous to "bots" in other chat applications, allowing different AI assistants to coexist in the same chat environment. Each persona can have unique behavior, models, and capabilities.

Adding a New Persona via Entry Points

To create and register a custom AI persona:

1. Create Your Persona Class

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

# Path to avatar file in your package
AVATAR_PATH = os.path.join(os.path.dirname(__file__), "assets", "avatar.svg")


class MyCustomPersona(BasePersona):
    @property
    def defaults(self):
        return PersonaDefaults(
            name="MyPersona",
            description="A helpful custom assistant",
            avatar_path=AVATAR_PATH,  # Absolute path to avatar file
            system_prompt="You are a helpful assistant specialized in...",
        )

    async def process_message(self, message: Message):
        # Your custom logic here
        response = f"Hello! You said: {message.body}"
        self.send_message(response)

Avatar Path: The avatar_path should be an absolute path to an image file (SVG, PNG, or JPG) within your package. The avatar will be automatically served at /api/ai/avatars/{filename}. If multiple personas use the same filename, the first one found will be served.

2. Register via Entry Points

Add to your package's pyproject.toml:

[project.entry-points."jupyter_ai.personas"]
my-custom-persona = "my_package.personas:MyCustomPersona"

3. Install and Restart

pip install your-package
# Restart JupyterLab to load the new persona

Your persona will automatically appear in Jupyter AI chats and can be @-mentioned by name.

Loading Personas from .jupyter Directory

For development and local customization, personas can be loaded from the .jupyter/personas/ directory:

Directory Structure

.jupyter/
└── personas/
    ├── my_custom_persona.py
    ├── research_assistant.py
    └── debug_helper.py

File Requirements

• Place Python files in .jupyter/personas/ (not directly in .jupyter/)
• Filename must contain "persona" (case-insensitive)
• Cannot start with _ or . (treated as private/hidden)
• Must contain a class inheriting from BasePersona

Example Local Persona

File: .jupyter/personas/my_persona.py

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

# Path to avatar file (in same directory as persona file)
AVATAR_PATH = os.path.join(os.path.dirname(__file__), "avatar.svg")


class MyLocalPersona(BasePersona):
    @property
    def defaults(self):
        return PersonaDefaults(
            name="Local Dev Assistant",
            description="A persona for local development",
            avatar_path=AVATAR_PATH,
            system_prompt="You help with local development tasks.",
        )

    async def process_message(self, message: Message):
        self.send_message(f"Local persona received: Hello world!")

Note: Place your avatar file (e.g., avatar.svg) in the same directory as your persona file.

Refreshing Personas

Use the /refresh-personas slash command in any chat to reload personas without restarting JupyterLab:

/refresh-personas

This allows for iterative development - modify your local persona files and refresh to see changes immediately.

Development install:

micromamba install uv jupyterlab nodejs=22
jlpm
jlpm dev:install

Example for connecting persona to Ollama

install Ollama and download model:

ollama pull deepseek-coder:6.7b
pip install ollama

persona code:

from jupyter_ai_persona_manager import BasePersona, PersonaDefaults
from jupyterlab_chat.models import Message
import os
from ollama import chat
from ollama import ChatResponse

# Path to avatar file (in same directory as persona file)
AVATAR_PATH = os.path.join(os.path.dirname(__file__), "avatar.svg")


class MyLocalPersona(BasePersona):
    @property
    def defaults(self):
        return PersonaDefaults(
            name="NeurodeskAI",
            description="A persona for local development",
            avatar_path=AVATAR_PATH,
            system_prompt="You help with local development tasks.",
        )

    async def process_message(self, message: Message):
        response: ChatResponse = chat(model='deepseek-coder:6.7b', messages=[
          {
            'role': 'user',
            'content': message.body,
          },
        ])
        print(response['message']['content'])
        # or access fields directly from the response object
        self.send_message(response.message.content)

Requirements

• JupyterLab >= 4.0.0

Install

To install the extension, execute:

pip install jupyter_ai_persona_manager

Uninstall

To remove the extension, execute:

pip uninstall jupyter_ai_persona_manager

Troubleshoot

If you are seeing the frontend extension, but it is not working, check that the server extension is enabled:

jupyter server extension list

If the server extension is installed and enabled, but you are not seeing the frontend extension, check the frontend extension is installed:

jupyter labextension list

Contributing

Development install

Note: You will need NodeJS to build the extension package.

The jlpm command is JupyterLab's pinned version of yarn that is installed with JupyterLab. You may use yarn or npm in lieu of jlpm below.

# Clone the repo to your local environment
# Change directory to the jupyter_ai_persona_manager directory
# Install package in development mode
pip install -e ".[test]"
# Link your development version of the extension with JupyterLab
jupyter labextension develop . --overwrite
# Server extension must be manually installed in develop mode
jupyter server extension enable jupyter_ai_persona_manager
# Rebuild extension Typescript source after making changes
jlpm build

You can watch the source directory and run JupyterLab at the same time in different terminals to watch for changes in the extension's source and automatically rebuild the extension.

# Watch the source directory in one terminal, automatically rebuilding when needed
jlpm watch
# Run JupyterLab in another terminal
jupyter lab

With the watch command running, every saved change will immediately be built locally and available in your running JupyterLab. Refresh JupyterLab to load the change in your browser (you may need to wait several seconds for the extension to be rebuilt).

By default, the jlpm build command generates the source maps for this extension to make it easier to debug using the browser dev tools. To also generate source maps for the JupyterLab core extensions, you can run the following command:

jupyter lab build --minimize=False

Development uninstall

# Server extension must be manually disabled in develop mode
jupyter server extension disable jupyter_ai_persona_manager
pip uninstall jupyter_ai_persona_manager

In development mode, you will also need to remove the symlink created by jupyter labextension develop command. To find its location, you can run jupyter labextension list to figure out where the labextensions folder is located. Then you can remove the symlink named @jupyter-ai/persona-manager within that folder.

Testing the extension

Server tests

This extension is using Pytest for Python code testing.

Install test dependencies (needed only once):

pip install -e ".[test]"
# Each time you install the Python package, you need to restore the front-end extension link
jupyter labextension develop . --overwrite

To execute them, run:

pytest -vv -r ap --cov jupyter_ai_persona_manager

Frontend tests

This extension is using Jest for JavaScript code testing.

To execute them, execute:

jlpm
jlpm test

Integration tests

This extension uses Playwright for the integration tests (aka user level tests). More precisely, the JupyterLab helper Galata is used to handle testing the extension in JupyterLab.

More information are provided within the ui-tests README.

Packaging the extension

See RELEASE