Tana Productivity: A Workflow for Notes, Tasks, and AI

A useful Tana productivity workflow has four parts: capture a note, identify the action it contains, review the actions you can actually do, and save what changed. Start with one real project and a small structure, then add complexity only when a repeated problem needs it.

Tana can be your working notebook while fenbs holds the agreed decisions and progress that different AI apps need to reuse. That distinction lets you think freely in notes without making every unfinished thought an instruction for the next assistant.

How do you set up a useful first workflow?

1. Choose one active project with an outcome you can describe in a sentence. A customer onboarding guide is easier to organize than a category called Work.
2. Create a project note with the goal, the source material, unresolved questions, and the next action. Keep those labels consistent across your first few projects.
3. Give actions a recognizable task structure. Include a project reference, an owner when relevant, and a completion condition.
4. Create a review point for unfinished actions. Open it at the start of work, select the next useful action, and return to the project note for its background.
5. End the session by recording the result and the next step. Move accepted project conclusions into the shared record when someone else needs them.

These are workflow choices, not a requirement to reproduce a complicated dashboard. For the onboarding example, the first action could be to identify three questions the guide must answer. You do not need a tag for every department before you can do that work.

What structure keeps notes connected to tasks?

Tana calls a structured tag a supertag: it gives a note a type and a reusable set of fields. Its supertag documentation↗ (opens in a new tab) explains how a task or project can use the same structure wherever you capture it. Begin with those two types rather than a large catalog.

• Project: outcome, current question, useful sources, and the next review point.
• Task: action, project, owner, and the evidence that will count as completion.
• Ordinary note: observations and thinking that do not yet need their own fields.

Use a project reference instead of repeating its name in slightly different forms. A note about the guide and a task to revise the guide should point to the same project. Reserve dates for real timing needs; an arbitrary date on everything makes a task list look urgent without making it useful.

For concrete action wording, see the AI task examples. Keep a task small enough that you can tell whether it happened, while leaving the research behind it in the project note.

How do you stop captured tasks from disappearing?

A capture system needs a reliable way back to the work. Tana search nodes↗ (opens in a new tab) display matching notes as references, and refresh when expanded. Use the query builder to collect unfinished tasks for your project, then test the result with a task you know should appear.

Check the opposite case too. Complete a sample task and confirm it leaves the unfinished view. A view that looks organized can still omit the work you need because one task has no project field or uses a different tag.

At your daily review, decide whether each visible action is ready, waiting, or no longer needed. If it is waiting, write what would unblock it. If a task says only “onboarding,” rewrite it as an observable action such as “Draft the password reset instructions and check them against the product.”

A personal filtered view and a shared work record solve different problems; the task manager versus task board guide explains the distinction. In fenbs, save the agreed state of shared work so a teammate does not have to interpret your private review list.

Where can AI help without creating more cleanup?

Use AI for a bounded transformation of material you have selected. Tana documents AI chat with notes as context↗ (opens in a new tab). Choose the relevant project notes, ask a narrow question, and compare the answer with the original material before making it part of the project record.

Read these selected onboarding notes.
List proposed actions with the source note for each.
Separate questions, suggestions, and decisions already made.
Do not invent an owner, deadline, approval, or completed check.
End with anything you could not establish from the notes.

Review each proposed action rather than accepting the list as a unit. Two notes might describe the same problem, while an emphatic sentence might be a customer request rather than an approved requirement. A human should decide which actions belong in the plan.

For a broader category comparison, use the AI productivity tools guide. This workflow is about making one notebook useful, not collecting another set of apps to maintain.

How do you review progress without reviewing every note?

Review by outcome. Ask what moved the project forward, what evidence supports that claim, and what still prevents delivery. A long list of captured notes is activity; a checked set of instructions is a result.

For the onboarding guide, keep the distinction between a drafted section, a reviewed section, and instructions tested against the product. Record which check happened and which did not. That makes the next session concrete even when another person takes over.

Remove fields you consistently ignore. If you never use a category to find or choose work, it may be an extra decision at capture time. The second-brain guides cover ways to organize reusable knowledge; your productivity setup should earn its complexity through actual reuse.

How to keep the shared project context with fenbs?

1. Create a free account, free for one person with no credit card, and add the project you are testing.
2. Add the active tasks and selected source links. Write the problem and the next action clearly enough for someone who has not read your notebook.
3. Record what a person approved in Decisions and rules. Keep unresolved choices open and say who needs to decide.
4. Connect a supported assistant using Integrations. MCP, a standard connection to outside tools, lets the assistant read and update the permitted work record after browser sign-in.
5. Ask the assistant to read the project context before working and save Where we left off afterward, including completed checks, uncertainty, and the next step.

fenbs does not import your Tana notebook. It is not the right fit for replacing a personal graph of every idea; keep Tana when linked note exploration is the work. Use the shared record when accepted conclusions need to survive changes of person or AI app, as explained in how fenbs works.