Deployable static outputs

Artifacts an agent can read without bloating the skill.

These files are built locally from the mirrored course and shipped as static assets. The site is just the browseable shell around them.

Core machine-readable files

llms.txt

Front-door manifest for agents. Links to the best next files instead of embedding content in the skill.

openapi.json

Read-only endpoint map for agents that prefer API-shaped navigation over raw file lists.

task-paths.md

Curated starting points for timing, performance, accessibility, interaction design, and review work.

animation-all.md

Full combined Markdown context for broad reads and offline prompting.

llms-ctx.txt

Smaller quick-start context file optimized for first-pass routing.

Local query examples

sqlite3 public/db/animation-content.sqlite \
'SELECT lesson_title, heading, url_path
 FROM sections_fts
 WHERE sections_fts MATCH ''transform origin''
 LIMIT 5;'
bun run search -- "transform origin" --limit 5
rg -n "prefers-reduced-motion|scale\\(0\\)|ease-out" public/md

Lesson Markdown

51 generated files