Roam Research Review: Linked Notes for Real Work

Roam Research is a strong fit when you think in outlines and want ideas connected across notes, rather than stored only in separate documents. Its value depends on whether those connections help you retrieve evidence, develop an argument, or continue real work. A tool such as fenbs can sit alongside personal notes when the next problem is giving connected AI assistants a shared record of tasks, decisions, and progress.

This is a review of the documented workflow and its tradeoffs, rather than a claim of hands-on performance testing. The useful question is whether Roam’s way of organizing information matches your daily habits. A short trial with real notes will tell you more than a feature count.

For practical setup instructions, see the guides for students and project work; for a wider shortlist, see Roam alternatives.

What makes Roam’s note structure useful?

Roam organizes pages as outlines of blocks. A block is a bullet that can have details nested beneath it. You can connect named pages and refer to individual blocks, as described in the official syntax guide↗ (opens in a new tab). This makes the small passage, rather than only the whole document, a useful unit of work.

Suppose you research how a neighborhood’s transit access changed. You might have separate notes for an interview, a planning report, and your own analysis. Linking the same concept across them gives you a route through the evidence without forcing all three into one long document. The benefit comes from useful relationships, not the number of links.

This structure favors people who naturally write short points and develop them over time. If you prefer a polished page with a settled beginning and end, an outline may feel like an intermediate stage you still need to convert. That is a working-style tradeoff, not proof that either approach is more serious.

For the broader practice behind the tool, see our guide to an AI second brain. The tool can store connections; you still choose what deserves to become reusable knowledge.

Where do the connections save effort, and where do they add it?

Connections help when the same idea matters in several places. Instead of rewriting a source explanation for every project, point back to the relevant passage and add your interpretation. Keep enough surrounding context to understand what that passage actually supports. A link alone does not establish that two claims agree.

Roam’s reference documentation↗ (opens in a new tab) distinguishes a page link, a block reference, and an embed. They have different purposes. Learn the difference before building a large collection around them, especially if you intend to reuse passages in several project outlines.

The maintenance cost is deciding what to name, what to connect, and which statement remains current. A graph full of vague pages can be harder to use than a modest folder of clear notes. A practical review should therefore include a delayed retrieval test: return after several days and find the evidence behind an important conclusion.

• A useful link explains a recurring concept, source, person, or project.
• A useful note distinguishes the source’s claim from your interpretation.
• A useful project page identifies current work without requiring a tour of the whole graph.
• A useful review removes uncertainty about what remains valid and what has changed.

Can AI assistants work with a Roam graph?

Yes. Roam publishes an official MCP and command-line tool↗ (opens in a new tab), allowing supported assistants to read, write, search, and query a graph. MCP means a standard connection through which an AI app can use an outside tool. Roam is therefore more than a passive source you can only copy and paste.

That access deserves an explicit plan. Test with disposable content, check the permissions you are granting, and decide which operations need your review. Ask the assistant to explain a proposed change before editing important source notes. A successful connection is not evidence that the assistant understood your preferred structure or the meaning of a reference.

Separate access to notes from instructions about the work. An assistant may retrieve a design discussion and still need to know whether the design was approved. If several AI apps help with a project, define which record holds the current decisions and where progress gets saved. That is the practical difference between a note archive and a shared AI brain.

What should a team test before relying on it together?

Start with a small shared-work exercise using the access options available in your Roam setup. Ask another person to find the current goal, identify an open question, and leave a response. Check the current vendor help and your actual settings before putting sensitive team information in the graph.

Then test a handoff with someone who did not attend the last meeting. Can they distinguish the original suggestion from the approved decision? Can they identify the responsible person and the next action? These checks reveal whether your conventions work for others, rather than merely feeling familiar to the person who built them.

Treat shared understanding and shared editing as separate needs. People can contribute to a collection without agreeing which statement is authoritative. Assign someone to maintain current project summaries, and make the distinction between research, proposals, and settled instructions visible in the text.

fenbs addresses that work record with projects, tasks, Decisions and rules, and Where we left off; how it works shows the structure. Team members have roles, and changes appear in History. Your connected notes can stay in Roam while selected work commitments have one maintained home.

How portable is a workflow built around linked blocks?

Judge portability by what another tool or person can understand after the content leaves its original setting. Plain words travel more easily than interactive relationships. For an important project, make a small sample that includes a source note, a reference to a passage, an open question, and a current decision. Check that the meaning survives when you read it elsewhere.

Roam’s official connection includes a tool that returns page content as Markdown, a plain-text format with lightweight formatting. Its architecture documentation↗ (opens in a new tab) also distinguishes desktop and hosted connection capabilities. Do not assume every connection exposes the same features or that readable text reproduces the entire interactive graph.

For a human handoff, write the relevant conclusion in a complete sentence and retain the source location. For a move to another AI app, carry the approved instructions and current progress as well. Our portable AI memory guide explains why those selected facts matter more than preserving a previous conversation word for word.

How to do this with fenbs

1. Create a free account, free for one person with no credit card, and add a project already represented in your notes.
2. Write its current goal and add the active tasks. Include selected source links and findings from Roam yourself, keeping exploratory notes in the original collection.
3. Record the choices people have actually approved in Decisions and rules. Keep unresolved questions open and identify where the next session should begin.
4. Connect a supported AI app through Integrations. Ask it to read the project context, explain its next action, and save progress before stopping.

Roam is worth considering when linked outlines help you think and retrieve evidence with less repetition. fenbs is not the right fit if all you need is a personal network of research notes. Add it when coordinating shared work and keeping AI sessions aligned becomes a separate, recurring need.