Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Design hands-on technical nonfiction chapters as runnable labs, worked examples, exercises, checkpoints, troubleshooting paths, and reader outcomes. Use when turning technical book sections, tutorials, self-hosting chapters, DevOps guides, programming lessons, or infrastructure explanations into practical labs that readers can complete and verify.
.claude/skills/hashgraph-online-technical-book-lab-design/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -22% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 54% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 44% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 73% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 43% | 0% |
A technical book lab is not a pile of commands. It is a guided reader transformation with a clear starting state, safe path, visible progress, expected output, and exit criteria. The reader should know what they built, why it works, how to verify it, and what to do when it fails.
Use this skill to design or improve:
| Need | Read | |------|------| | Core lab concepts and terminology | references/core/knowledge.md | | Lab design rules | references/core/rules.md | | Before/after examples | references/core/examples.md | | Lab review checklist | references/core/checklist.md | | Step-by-step lab design | workflows/design-runnable-lab.md |
State what the reader will be able to do, decide, deploy, debug, or explain after the lab.
If the outcome is only "learn about X," sharpen it until there is visible proof of progress.
List what the reader already has and knows:
Move missing prerequisites earlier, compress them, or link them to companion material.
Structure the lab as small checkpoints:
Technical readers do not only need the happy path.
End every lab with:
When designing or reviewing a lab, return:
Make the lab useful to a real reader in a real environment. Prefer concrete outputs, checkpoints, and failure handling over abstract explanation or untested command sequences.
Other measured skills in the registry, with their headline benchmark lift.