Scaffold Numbered Course Exercises

What it does

The skill turns a course outline into numbered sections and exercises under exercises/. Each exercise contains a problem, solution or explainer variant; unspecified stubs default to explainer. Variant readmes must have real content, and code variants need a meaningful main.ts. The publisher’s ai-hero-cli linter checks structure and links. Existing exercises are moved with git mv when renumbering. A structurally valid scaffold is a starting point for teaching material, not proof that a lesson is correct or complete.

When to use it

• Create the first directory structure from an approved course outline.
• Add a section containing matched problem and solution variants.
• Renumber exercises while keeping links and Git history understandable.

Step-by-step workflow

1. Read the outline and assign section and exercise numbers.
2. Choose variants explicitly, defaulting an unspecified stub to explainer.
3. Create each directory and a nonempty descriptive readme.
4. Add code only for variants that need it and update links when moving files.
5. Run the existing course linter, fix structural failures and review authorized commit scope.

Example request

Use scaffold-exercises for a section on retrieval evaluation. Create an introductory explainer and one problem/solution pair comparing two retrieval results. Check our current numbering before choosing prefixes. Add real titles and descriptions instead of empty placeholder files. Use the existing ai-hero-cli lint command if available; report a missing tool instead of installing it. Do not commit unrelated files.

Worked example

A section numbered 04 might contain 04.01-introduction with an explainer and 04.02-compare-results with separate problem and solution directories. The problem readme should tell the learner what to inspect; the solution readme should explain the expected comparison. A readme-only introduction does not need a fake TypeScript file to satisfy a presumed template. If an exercise moves, incoming relative links need checking as well as the renamed directory.

What you get

A consistent exercise scaffold with meaningful readmes, required code entry points and a structural validation result.

How to evaluate the result

• Check section and exercise prefixes match.
• Verify readmes are nonempty and links resolve.
• Confirm every requested variant exists without unrelated placeholder files.

What you need

A course plan, the repository’s current numbering and the publisher-specific ai-hero-cli environment for its exact lint command. Plain scaffolding can be reviewed separately if that tool is absent.

Setup and activation

The publisher documents plugin setup for Claude Code, Codex and Copilot, and manual Agent Skills setup for Gemini CLI and other clients. Choose one distribution method per client to avoid duplicate skills. The engineering collection expects project-specific tracker, triage-label and document-location settings from setup-matt-pocock-skills. Keep referenced companion skills and files available. This catalogue review does not install or run them; scan the exact source before any later installation. Client tools, permissions and subagent support still need separate validation.

Limits and considerations

The lint rules are specific to the source’s course format. They do not establish instructional quality, correctness or accessibility. The upstream description includes a commit; publishing that commit needs the user’s actual authorization.

Related workflows

• Use teach to develop and evaluate the learning activities.
• Use writing-shape for shaping an explanatory article from existing notes.

License and reuse

MIT at repository level (Copyright 2026 Matt Pocock). Preserve the copyright and permission notice when redistributing substantial source material; inspect any separately licensed dependencies or assets.

Further reading

• Reviewed SKILL.md (immutable commit)
• Publisher README: distributions and compatibility
• Reviewed MIT license