A few practical questions.
From your first map to keeping the whole team in sync.
Mapping your system
Does Nodeberg discover my architecture automatically?
You define the relationships in the visual editor or Nodeberg YAML. Your CI pipeline can publish that YAML to keep the map updated as your architecture documentation changes. Nodeberg does not scan repositories, discover infrastructure, or monitor live traffic.
What can I see when I select a service?
Its documented direct callers and call targets over any protocol, the events it produces and consumes, and the services connected through those events. Upstream, Downstream, and Relevant only let you focus the view. This is a local dependency view, not a guarantee of every transitive runtime impact.
Can I describe a change in words?
Yes. Choose Update map with AI above the services in the sidebar, write something like “the API now calls Twilio for OTP”, and the assistant proposes the edits: the services you mean, a new service if the map lacks one, and the call between them. On an empty map, Describe your system does the same for a whole small system. Requests are limited to 560 characters, about two tweets. You review the list and apply it, or discard it; nothing is saved until you apply, and the apply is a normal revision you can restore. Say “we removed Job Feed service” and it proposes removing that service and its incoming and outgoing calls, without transferring anything to another service. Events other services still use stay. You can also tell it that several services are really one, like “the API, the worker and the scheduler are a single service”, and it merges them into one that keeps what they had. Apart from removing and merging services, the assistant only adds. To work out the edits it sends your text and the map’s service names and IDs, event names and call protocols to an AI provider’s API, OpenAI or Anthropic depending on how Nodeberg is configured. Nothing is sent unless someone uses it, and descriptions and links are never sent.
What is a map?
One architecture document containing services, the calls between them (HTTP, gRPC, or whatever protocol they use), events, and their layout. Use separate maps for different systems, projects, or architecture proposals.
Can I take my documentation with me?
Yes. Export a Nodeberg YAML document with your services, connections, and layout. You can import it again or keep a copy alongside your project documentation. Importing opens a draft for review before you apply it. You can also download the map as a PNG or SVG image for design docs, wikis and slides.
Syncing from CI
How does syncing from CI work?
A workspace manager creates a sync source in Manage workspace → Developers, then adds its upload token to the pipeline’s secrets. Keep your service definitions in Nodeberg YAML and publish them from CI. Copyable GitHub Actions, GitLab CI, and curl examples are built into the app. Each upload replaces that source’s contribution to the map, including removing services omitted from the YAML.
Can several repositories contribute to one map?
Yes. Give each publishing pipeline its own source. A source can define one service or several, and reference services from other sources on the same map. Shared service IDs connect those contributions. Nodeberg prevents one source from overwriting another source’s services; taking over existing manual services requires a manager’s review and approval.
Can I still edit a synced map?
Yes. Arrange the canvas and edit manual services as usual. Synced service definitions are read-only in Nodeberg: change their source YAML and publish again. Disconnecting a source revokes its upload token and makes its services manually editable, while keeping the services and sync history.
How do I know what changed?
Each source shows its latest attempt and last successful sync. Open its history to inspect upload outcomes, affected services, submitted YAML, and any commit, branch, or pipeline references supplied by your CI job. When a newer map revision is available, Nodeberg prompts you to reload and asks before discarding unsaved changes.
Does Nodeberg need access to my repositories?
No repository credentials are needed. Your pipeline sends the YAML and any commit, branch, or pipeline references you choose to include. Its token only authorizes uploads to one sync source, and a workspace manager can rotate or revoke it.
Teams and pricing
Is the price per person?
No. Plans are priced per workspace, with a people limit on each. Plus costs $29.99/month or $290/year as a limited founding price for early teams, locked in while you stay subscribed (the regular price is $49.99/month or $490/year), billed upfront, in USD, with 10 maps, up to 25 people and 10 editors. Pro costs $99.99/month or $999/year, with 30 maps, up to 100 people and 25 editors. Prices exclude applicable taxes.
Who counts as an editor?
A person who can change maps anywhere in your workspace counts once as an editor, even if they edit several maps. Owners and administrators count as editors. Pipeline tokens and people only committing YAML in Git do not count. Everyone in the workspace, editors and viewers alike, also counts toward its people limit, and editors have their own limit within it.
Can multiple people edit the same map?
Yes, with editing access. Saved changes are shared, and conflicting saves are rejected so an older draft cannot silently replace a newer version. Editing is not synchronized live between users.
What does the free plan include?
Free includes 1 map per account and up to 3 people, with the same mapping and CI sync features as paid plans. Plus raises that to 10 maps and 25 people, and Pro to 30 maps and 100 people.
What happens if I cancel or downgrade?
Everything is kept and stays viewable, and export always works. Maps beyond your plan become read-only until you are within its limits again, or you upgrade. Sync uploads to a read-only map are rejected until then.
Ready to put your system on the map?
Create your first map