A Todo list that understands the project behind it
The new optional Todo skill keeps follow-ups attached to real projects, whether you are shipping software, finishing a book, or trying to make a personal goal concrete.
Most Todo lists answer one question: what do I need to do next?
That is useful, but it leaves out the question that makes the next action meaningful: next for what?
“Write the onboarding page” belongs to a product. “Finish chapter four” belongs to a book. “Walk three times this week” belongs to a health goal. Without that project around it, a Todo becomes another orphaned reminder competing for attention.
The new optional Empirical Todo skill gives an agent a better place to put those follow-ups.
A project, not a pile
When you tell an agent, “Add a Todo to update the Empirical plugin README,” the useful part is not just the sentence. It is the project relationship: this belongs to the Empirical memory plugin, it is still open, and the next action is to explain the install choices clearly.
Later, you can ask what remains for that project and get actionable work instead of a stream of unrelated reminders.
The same pattern works outside software.
You might say, “I just finished 12 Rules for Life. Add the follow-ups I want to try in my own life.” The agent can turn that conversation into a few concrete actions attached to a personal project, rather than pretending every idea is a permanent fact about you.
Or you might say, “Track my plan to get back into regular workouts.” The project is the goal. The Todos are the small actions that move it forward. The memory layer can hold the durable preference or decision; Todo holds the unfinished work.
Different from memory, history, and scar tissue
This separation is the point.
- Memory keeps durable context: decisions, preferences, and facts worth recalling.
- Work History records what was completed and verified.
- Scar Tissue records what went wrong and how to avoid repeating it.
- Todo keeps an outcome moving toward completion.
They can belong to the same project without becoming the same kind of note. That makes it easier for an agent to decide whether it should recall context, record a completed handoff, warn about a known sharp edge, or simply show the next action.
Todo is optional. If you do not want an agent managing follow-ups, you do not install it. If you do install it, the agent creates or updates Todos when you ask for that behavior; it does not quietly turn every thought into a task.
A useful second brain is allowed to be selective
The promise of a second brain is not that everything gets saved. It is that the right things stay connected.
That is why project scope matters. A book can generate personal experiments. A product can generate launch work. A health goal can generate weekly actions. Each project gives the agent enough shape to help without flattening your life into one giant inbox.
Read the Todo plugin guide, or start with Plain Markdown if your projects already live in portable files. The broader idea is in Your AI needs your second brain.