Loom Learning Graph Skill

This skill guides agents on how to use the Loom plugin to build and expand a learning graph over time.

Purpose

• Help users navigate learning paths (e.g., Nix, German) through graph nodes.
• Expand the graph organically as new knowledge appears.
• Keep knowledge warm with spaced repetition reviews.

Core Model

• Path: a learning track (e.g., nix, german).
• Node: atomic learning unit with prerequisites, unlocks, and review scheduling.
• Context: real-world capture that should seed new nodes or enrich existing ones.

What a Node Contains

• A short explanation in your own words
• A concrete example or mini exercise
• Links to prerequisite or follow-up nodes by id
• A “check yourself” prompt (question or task)

When to Create Nodes

• When a user says “I want to learn X”, create a path for X and seed 1-3 starter nodes.
• When a user mentions a new concept, term, or confusion point, create a node.
• When a context capture reveals missing scaffolding, add prerequisites.
• When a node reaches mastery, add or unlock a next node that builds on it.
• When a session reveals gaps, insert a bridging node.

How to Expand the Graph

• Prefer small, focused nodes over large, vague ones.
• Use prerequisites to gate advanced topics.
• Add unlocks to highlight natural next steps.
• Always link nodes by id in body text (wikilinks are fine).

Workflow (Agent)

1. Capture context when the user mentions a real situation.
2. Create or update nodes from the context.
3. Unlock nodes when prerequisites are mastered.
4. Recommend next using learn next and prompt review with learn review.

Tooling

• learn_add_node to add nodes.
• learn_capture to capture contexts.
• learn_next to pick the next node.
• learn_review to schedule repetition.
• learn_query to search the library.

Defaults

• Nodes are Markdown (.md).
• Spaced repetition intervals default to [1, 3, 7, 14, 30, 60, 120, 240] days.

Example Triggers

• “I read about Nix flakes today” → add node nix/flakes with prereqs.
• “I got confused ordering coffee in German” → capture context and add node german/ordering-coffee.