Sage Agent — The AI-Powered Assistant for Jupyter Notebooks
What is Sage Agent?
Sage is an AI-native notebook assistant that supercharges your existing Jupyter workflows.
Built by leading AI and quant researchers from YC, Harvard, MIT, and Goldman Sachs, Sage brings real-time, context-aware assistance directly into JupyterLab.
Use natural language to clean data, write analysis code, debug errors, explore dataframes, and build models—faster and with fewer mistakes.
No hallucinated code. No context switching. Just faster insights.
Why Use Sage Agent in Jupyter?
Whether you’re a quant, data scientist, or analyst living in notebooks, Sage helps you:
✅ Clean and transform messy data in seconds
✅ Visualize trends, rollups, and anomalies from a prompt
✅ Connect your custom databases in one click and easily explore from notebooks
✅ Generate runnable Python or SQL that fits your current cell + variable context
✅ Auto-detect schema changes and debug downstream errors
✅ Stay private: run entirely local-first or in your own secure VPC
✅ Extend JupyterLab without changing how you work
Perfect For:
- Data scientists cleaning huge CSVs
- Quant researchers testing ML pipelines
- Product and analytics teams tired of building dashboards and flaky notebooks
- Anyone tired of LLM tools that break their code
Installation
📦 Requirements
- JupyterLab >= 4.0.0
- NodeJS (for development)
🧠 Install Sage Agent:
pip install jupyterlab sage_agent_internal
❌ Uninstall:
pip uninstall sage_agent_internal
How to Get Started
To unlock full functionality, you’ll need Sage API credentials.
👉 Request your API key or email us at contact@sagebook.ai
Why Sage
- ✅ Context-aware code gen: understands variables, dataframes, imports, and prior cells
- ✅ AI that fixes schema issues and silent join bugs
- ✅ Inline review + diffs before you run any code
- ✅ Visualizations via natural language (matplotlib, plotly, seaborn supported)
- ✅ BYO LLM: Anthropic, OpenAI, vLLM, Ollama, or HF endpoints
- ✅ Built to run in air-gapped / enterprise environments
Local Development Instructions
To contribute or develop locally:
# Clone the repo and enter the directory
git clone https://github.com/sagebook/sage_agent_internal.git
cd sage_agent_internal
# Install in editable mode
pip install -e "."
# Link extension to JupyterLab
jupyter labextension develop . --overwrite
# Rebuild on changes
jlpm build
For auto-rebuild while editing:
# Watch source
jlpm watch
# Run JupyterLab in parallel
jupyter lab
Uninstall in Dev Mode
pip uninstall sage_agent_internal
# Then manually remove labextension symlink from JupyterLab extensions dir.
Want to See Sage in Action?
🎥 Try the demo notebook or explore at https://sagebook.ai
Built for teams working with sensitive data:
- Zero data retention by default
- Optional BYO keys for Claude, OpenAI, or local models
- Notebook-specific controls for what the model can “see”
- Fine-grained telemetry settings
Contact
Questions? Ideas?
Email: fahim@sagebook.ai
Website: https://sagebook.ai
AI Jupyter Notebook, JupyterLab Extension, Jupyter Assistant, Data Science Assistant, Jupyter LLM, AI code generation, dataframe cleaning, Jupyter AI, Sagebook, Sage Agent, AI for dataframes, Jupyter SQL assistant, notebook extension
Metadata
Release files for sage-agent-internal 0.5.14
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| sage_agent_internal-0.5.14.tar.gz | 1.7 MB | Without provenance | Details for sage_agent_internal-0.5.14.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| sage_agent_internal-0.5.14-py3-none-any.whl | Python 3 | none | any | Details for sage_agent_internal-0.5.14-py3-none-any.whl |
Total release size: 2.6 MB
Release files / sage_agent_internal-0.5.14.tar.gz
| Download URL | sage_agent_internal-0.5.14.tar.gz |
|---|---|
| Size | 1.7 MB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
2b617e09828f711ed561e20f4b9213765084a1ebbb08be67b1ead1c9432ced5b
|
|
BLAKE2b-256 checksum How to use checksums |
11d31aefa8ad027a357f904048a63aa94759a3d986974ac39a9534687df06ded
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.13.7
|
Release files / sage_agent_internal-0.5.14-py3-none-any.whl
| Download URL | sage_agent_internal-0.5.14-py3-none-any.whl |
|---|---|
| Size | 961.7 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
d82303e3985af5947e2583411adcedd37749b046deee172365a712ca16af671c
|
|
BLAKE2b-256 checksum How to use checksums |
b7ba169854d37f60f2943f883d974a12f8e0a84ab3a60a8ef9ff47e13afe00ba
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.13.7
|
Security for sage-agent-internal 0.5.14
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.