illustrate

# Illustrate Turn one confirmed topic into a reopenable interactive explanation. ## Intake Read [learning context](../learning-context.md). Use a learning brief supplied by `/learn` as the source; otherwise resolve the current conversation plus any argument and confirm one topic. In both cases, recall relevant prior knowledge at invocation and refresh the brief before illustrating. Use recalled knowledge only to choose vocabulary, depth, and which connections need emphasis. Continue without personalization after reporting a recall failure. ## Artifact Create a self-contained HTML artifact under: ```text ~/.agentic/artifacts/illustrations/<repo>/YYYY-MM-DD-<topic-slug>.html ``` Append `-2`, `-3`, and so on when the name already exists. Resolve `<repo>` from the git root basename, then the current directory basename, then `no-repo`. The artifact must make these relationships visible: - the concept's central mental model, - important terms and their plain-language meanings, - dependencies, connections, or sequence, - relevant examples or source excerpts. Include meaningful interaction such as selecting nodes to reveal explanations, tracing a flow, switching views, or expanding examples. A static image with hover labels is insufficient. Keep all CSS and JavaScript inside the HTML so it can be reopened without a build step or network connection. Show or open the artifact with the environment's artifact-preview capability when available, and always report its absolute path. ## Boundary The artifact explains. It does not quiz, grade, infer mastery, or write to agent memory. At completion, append the configured `illustrate` suggestions from [skill connections](../../../docs/skill-connections.md).