Second Brain

Turn scattered raw material into a living knowledge system that is easy to retrieve, update, audit, reuse, and publish from.

This skill is storage-agnostic. Prefer plain folders and markdown or text files unless the user already has an existing tool, database, or knowledge system. Do not assume Obsidian, a plugin ecosystem, or any specific app. If the user already has a workable system, map the workflow onto it instead of forcing a migration.

When the user already works in an Obsidian vault or wants compatibility with one, keep the storage contract portable and load references/obsidian_adapter.md. That reference adapts the second-brain model to Obsidian-style markdown, views, canvas files, CLI helpers, and clean web capture without making any of them required.

One useful mental model is:

• raw sources are collected into a durable source layer
• the assistant incrementally compiles those sources into a maintained markdown wiki
• the user interacts with the compiled knowledge through an editor, file browser, search UI, CLI tools, or another frontend
• answers, reports, slides, plots, or scripts generated from the wiki are treated as artifacts that can themselves be filed back into the system

The goal is not just to store notes. The goal is to make the knowledge base compound: