Second Brain Project

Set up a project

The knowledge layer answers what you know. The project layer answers what you are doing this week. Merging them fills a wiki with dead task lists and buries working files under conceptual notes.

Core rule

One goal per project, stated in its own CLAUDE.md. A project with three goals is three projects.

Workflow

1. Interview if needed. What this produces, for whom, by when, and what the

agent's role is. Vague answers produce a file that makes every future session worse.

1. Create the folders: Inputs/, Process/, Outputs/, Feedback/.
2. Write the project CLAUDE.md: what this is, the one goal with a date,

your role, working rules, current state. Name what the agent should not do here as well.

1. Link the knowledge layer. Find concept pages the project depends on and

reference them from the project file. Copy into Inputs/ only the few that genuinely matter.

1. Register it in the root CLAUDE.md project list.

Output format

Project: 
Goal: 
Folders created: Inputs, Process, Outputs, Feedback
Concept pages linked: 
Agent role: 

Calibration

Feedback/ is the folder people skip and the one that makes the system improve. If the user has no idea what would go in it, that is worth surfacing: it usually means the goal is not measurable yet.

Do not create a project for work that finishes this week. That is a task.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: undefined-ui
• Source: undefined-ui/second-brain-os
• License: MIT
• Homepage: https://x.com/undefinedKi/status/2068306794116501544

Install and usage instructions live in the source repository linked above.