concept-graph_skill

Overview

This skill builds a concept graph (nodes + prerequisite edges) from a tutorial spec and saves it as outline/concept_graph.yml. It produces a readable DAG that captures concepts implied by objectives and the running example. The output is focused on structure only, not long explanatory prose.

How this skill works

The skill reads output/TUTORIAL_SPEC.md, extracts and normalizes implied concepts into nodes with stable ids, and infers prerequisite edges. It verifies the graph is acyclic and writes outline/concept_graph.yml using a minimal YAML schema (nodes: id, title, summary; edges: from, to). It enforces node coverage of learning objectives and preferred node counts.

When to use it

• You have a completed output/TUTORIAL_SPEC.md and need to plan module order (C2 structure stage).
• Preparing to convert tutorial structure into prose and lesson modules.
• Designing assessment or example progression that depends on prerequisite ordering.
• Validating that learning objectives map to concrete teachable concepts.

Best practices

• Keep node count between ~10–30 for a medium tutorial to balance granularity and manageability.
• Choose stable, short ids (snake_case) and specific titles; avoid vague labels like “misc”.
• Model intermediate prerequisites explicitly to avoid overly linear chains.
• Run cycle detection after adding edges; split or refine concepts if cycles appear.
• Keep summaries concise (one-line) describing the concept scope and why it matters to the tutorial.

Example use cases

• Turn tutorial objectives and a running example into a prerequisite DAG for module ordering.
• Identify missing foundational concepts before writing lesson prose.
• Generate a concept-driven outline that maps to assessments and code examples.
• Refine curriculum structure by splitting coarse concepts into teachable subunits.