Potential Links
The most valuable connections are often the ones you haven't made yet. When you open a node, Jotaid surfaces missing links in three distinct layers, from certain to speculative.
The Three Layers
Open a node and scroll its detail view:
- Linked notes — notes that already point to this node with a real
[[link]]. These are done. - Unlinked mentions — notes that type this node's name in plain text but never linked it. Jotaid finds the literal text match; you decide whether to link.
- AI semantic suggestions — notes that are related in meaning even without sharing the exact words. This layer is AI-powered and runs on demand (it deliberately excludes the plain-text matches above, so you only see genuinely new ideas). Requires Pro and a configured key.
Understanding which layer a suggestion comes from tells you how much to trust it: layer 2 is a guaranteed word match, layer 3 is an interpretation.
Backfill
When many notes mention a node in plain text, linking them one by one is tedious. Backfill does it in one step: hover a node and its backfill action turns every plain-text mention of [[NodeName]] across the project into a real backlink.
Backfill is a literal text operation, not AI — it only touches exact name matches, so it's safe and predictable.
Semantic Pairs
Beyond a single node's backlinks, Jotaid's AI can suggest node-to-node relationships — pairs of concepts that seem related across your library but aren't linked yet. These appear as semantic pair suggestions in Node view.
When It Works Best
Potential links get richer as your library grows. A handful of notes won't reveal much; a few dozen linked notes give the AI enough signal to make meaningful suggestions.
Next Steps
- Co-occurrence Matrix — See the same relationships as a heatmap
- AI Workflows — How the AI suggestions are generated