A Second Brain Template for Work With AI

A useful second brain template gives sources, your interpretation, current project context, decisions, and next steps distinct places. Copy the structure below into a note or document, then fill it with one real project. fenbs can turn the work portion into a continuing record that your connected assistants read and update.

The template is deliberately small. It should help you resume useful work, not create another system you need to maintain before starting. Keep complete source documents in their appropriate home and put only the necessary references and conclusions into the working record.

What can you copy into your first project note?

Use this as a plain text outline rather than a form with required fields. Replace each prompt with a concrete fact, or mark the question as open. Leaving something visibly unknown is more useful than filling the space with an assistant’s plausible guess.

PROJECT
Name:
Goal and definition of finished:
Person responsible for decisions:

CURRENT CONTEXT
What is happening now:
What has been checked:
What remains uncertain:

SOURCES
Title and owner or author:
Original location and relevant section:
What this source supports:
Limitations or date checked:

MY EXPLANATION
What I think the evidence means:
What is interpretation rather than a source fact:
Related idea and reason for the connection:

DECISIONS AND RULES
What a person chose, who chose it, and why:
Instructions that continue to apply:
Open choices that need a person:

NEXT WORK
Next action and intended result:
What is waiting and on whom:
Where the next session should start:

Give the project a result you could recognize. “Research the market” is broad; “prepare a source-checked summary of three customer segments for review” tells you what the work should produce. The intended result also helps you decide whether a new source is relevant.

Treat the template as a working index. It points to evidence and explains what is happening. It does not need to contain the whole research collection, every meeting transcript, or the complete history of your conversations.

How do you keep sources separate from conclusions?

Write a source entry before writing the conclusion it supports. Include enough location detail that someone else can check the relevant passage. If a link opens a long report, name the section or page rather than making the next person repeat your search.

Then explain the conclusion in your own words and identify any inference you added. For example, a source might report interview answers; you may infer that a proposed service needs a different introduction. The interviews and your design recommendation are related, but they carry different kinds of authority.

Project: Revise the signup instructions
Source: Approved usability session notes, task 2
Source finding: Participants asked which account name to use
My interpretation: The label may not explain the expected input
Limitation: We have not tested replacement wording
Decision: Keep the current account flow; decision made by the project owner
Next action: Draft two clearer labels for review

Forte’s second brain introduction↗ (opens in a new tab) describes keeping useful learning and its sources. This template applies that general principle to a work record: a small conclusion remains attached to the material that lets you challenge or improve it.

Which parts should an AI assistant read first?

Give the assistant the instructions that govern the work, the current goal, and the next step before adding background reading. That sequence helps it understand why it is reading a source. Clearly label unverified notes so fluent language does not make them appear settled.

Ask it to repeat the important boundaries and unresolved questions before doing substantial work. In the design example, it should know that changing labels is under consideration while changing the account flow is outside the approved scope. Correct a mistaken restatement before asking for a draft.

The AI brain guide explains the continuing work record, while the AI second brain guide covers the broader knowledge collection. Use the template to keep those two jobs connected without making every saved note an instruction.

Only share material you have permission to share with that AI app. A source reference in a project note is not a grant of access, and a local file location does not mean an assistant can open it. Check what content the assistant actually received.

How do you adapt the template without adding clutter?

Keep the headings that change a real action. A solo designer may need a source location and review question; a research group may also need a clear owner for checking evidence. Add a field after you repeatedly miss the information, rather than anticipating every possible use.

You can paste the outline into your existing notes tool. Obsidian’s Templates documentation↗ (opens in a new tab) describes inserting predefined text into a note. That can save repeated typing, but inserting headings does not check their contents or keep a project current.

For a tool-specific layout, see the Obsidian guide or Notion guide. Keep their setup questions separate from this template’s job: making the meaning, evidence, and current state of one project easy to understand.

A repeatable operating procedure needs its own instructions and checks; the SOP template covers that intent. Your second brain entry can reference the procedure and explain where a particular run stopped, without copying the entire procedure into every project.

What should you save at the end of a session?

Replace vague progress with a specific account of the work. Say what you completed, what you checked, what the check showed, and what remains. “Drafted two labels; neither has been tested; owner review is next” is enough to prevent the next assistant from claiming the change is validated.

Keep decisions stable until a person changes them, and keep the current summary current. If a later decision replaces an earlier one, preserve the reason for the change. Otherwise a reader may find an old approval and mistake it for the current answer.

fenbs saves where work was left for the next session, so the template’s next-work section can become a maintained project record instead of a paragraph left behind in one chat.

How to do this with fenbs

1. Create a free account and add the project from your completed outline. fenbs is free for one person, no credit card.
2. Add the next actions as tasks and put the goal, checked facts, and open questions into AI context notes. Keep long reference documents in their existing home.
3. Record the source locations in Links and listings, and save choices that people actually made in Decisions and rules.
4. Connect a supported assistant, ask it to read the project context, then have it save task progress and Where we left off before the session ends.

These steps are a manual setup, not an automatic template import. fenbs is not the right fit if you only want a personal notebook for long notes or a long-form writing editor. Use it for continuing work, and find related organization guides in the second brain topic hub.

For your first check, open a fresh conversation and ask the connected assistant to explain the project’s next action and remaining uncertainty. Compare its answer with the source-backed record. Adjust unclear wording before expanding the system to more projects.