FAIR implementation on lesson level - #536

Adds a machine-readable metadata pipeline for the lesson and a "Reusing" page that surfaces it.

• metadata.yml (new) — single source of truth for the lesson's citation/schema metadata: title, abstract, authors, version, DOI, license, repo/website URLs, keywords, and the teaches field (learning outcomes lifted from content/index.rst).
• .github/workflows/generate-metadata.yml + .github/scripts/generate_metadata_files.py (new) — on push to main touching metadata.yml, regenerates CITATION.cff and a new bioschemas.yml from it and commits them back.
• .github/workflows/zenodo.yml + .github/scripts/publish_to_zenodo.py (new) — on a published GitHub release, creates a new Zenodo version under the lesson's existing concept DOI and publishes it with metadata from metadata.yml.
• content/reusing.md (new) — citation info (DOI, CITATION.cff pointer), a live-rendered metadata table, and the license section. content/index.rst's "Reusing" toctree entry now points here instead of the generic coderefinery.org/lessons/reusing/ page.
• content/lesson_metadata.py (new) — Sphinx extension adding a {lesson-metadata} directive that renders metadata.yml as an HTML table (used on the new page).
• content/conf.py — registered lesson_metadata in extensions and uncommented sys.path.insert(0, os.path.abspath('.')), both required for the new extension to import.