Practical overview
When it fits
Use Second Brain when you want to assemble and run an extensible agent on your own machine, connect a model, and add documented tools or packages for local-information and workflow tasks.
Who it is for
- Developers and technical users setting up a local agent framework.
- People who want an LLM assistant to use local files and workflow tools.
- Users extending an agent with packages or messaging interfaces.
Documented uses
- Connect a language model to a locally running agent.
- Use file tools, shell access, web search and subagents through the essentials bundle.
- Add capabilities such as searching personal documents or maintaining Markdown-based memory.
- Explore an agent interface through a web UI or terminal chat.
Requirements & limits
The README documents local installation and lists several model-connection routes. It specifies Python 3.11+, Git and, for the web UI, Node.js 20.19+ or 22.12+. The repository is identified as MIT-licensed.
- The web UI requires Node.js in addition to the Python and Git prerequisites.
- The user must choose and configure a model connection during /setup.
- The essentials bundle includes file tools, shell, web search and subagents.
- The knowledge-base package is described as a large download; package installation is a separate step.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Install Second Brain and connect a model
Set up the local application, start its web interface and reach the model-connection stage.
Before you start
- Python 3.11 or later
- Git
- Node.js 20.19+ or 22.12+ for the web UI
Steps
Clone the repository and create a Python virtual environment named .venv.
Activate the environment, install dependencies with pip install -r requirements.txt, then run npm install inside frame_ui.
Return to the repository root and run python main.py; keep the terminal open while using the web UI.
In the web UI or terminal, run /setup and follow the model-connection prompts.
Expected outcome
The application starts locally and its setup flow presents options for connecting a language model.
Guide limitations
- Model connection requires a provider route: the README lists Codex sign-in, an OpenRouter API key, or another provider, including Ollama or an OpenAI-compatible endpoint.
- The README says the web UI needs Second Brain to remain running in its terminal.
Add a capability package
Use the package manager to browse and install an additional documented capability.
Before you start
- A running Second Brain installation with model setup completed
Steps
Run /packages install, or open Settings in the web UI.
Browse the package store and choose a package, such as bundle_knowledgebase or bundle_memory.
Expected outcome
The selected package is installed through the documented package-store flow.
Guide limitations
- The README notes that bundle_knowledgebase involves a large download.
- The guide does not specify package-specific setup or usage steps.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 9 October 2026
- Original website title
- henrydaum/second-brain · Official homepage title · retained in its source language