
Open Notebook Research Assistant
Lets an agent manage Open Notebook research notebooks, sources, and cited chats via its REST API.
- Repository
- K-Dense-AI/scientific-agent-skills
- Stars
- 55,780
- Licence
- MIT
- Files
- 9 files
What it does
Open Notebook is a self-hosted alternative to NotebookLM for organizing research. With this skill installed, the agent knows how to create notebooks, add sources such as PDFs, web pages, audio, and documents, chat with those sources with citations, run searches, and write notes, by calling a running Open Notebook server. It also includes Python helper scripts and reference docs for setting up AI model providers.
Great for
- Researchers who already run or want to run a self-hosted Open Notebook instance and want an agent to drive it
- Developers automating literature or source workflows through the Open Notebook REST API
- Students comparing sources in a notebook with cited answers
Skip it if
- People without a running Open Notebook backend, worker, and Docker or equivalent setup, since the skill cannot work without them
- Sensitive material sent to cloud AI providers, because content goes to those providers even though the app is self-hosted
- Anyone expecting the skill to verify AI answers; the docs warn that ingestion and paid AI calls were not tested against a live instance
Try asking
- “Create a notebook called Methods review and add these three PDFs as sources”
- “Chat with my notebook sources and give cited answers about the reported experimental designs”
- “List my notebooks and show which sources failed processing”
- “Set up Open Notebook with Ollama as the local model provider”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 9 original files, the MIT licence and an attribution note.
open-notebook.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip open-notebook.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip open-notebook.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Set OPEN_NOTEBOOK_ENCRYPTION_KEY to a real secret in the Compose file before first start, since the upstream placeholder is literal and the key must persist across restarts
- Set OPEN_NOTEBOOK_PASSWORD if the instance is reachable from other machines, and check sources reach processed status before asking for analysis
- Use local providers such as Ollama if documents must stay on your own hardware
What's inside
- references/api_reference.md
- references/architecture.md
- references/configuration.md
- references/examples.md
- scripts/_common.py
- scripts/chat_interaction.py
- scripts/notebook_management.py
- scripts/source_ingestion.py
- SKILL.md
Redistributed unmodified from K-Dense-AI/scientific-agent-skills under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.