Skip to content
Back to skills

Puzzle Design

ASecurity

Use when a puzzle needs explicit legal verbs, properties, solution policy, reset behavior and evidence without assuming one room, one answer or one teaching sequence.

  • 11 stars
  • 0 votes
  • 0 copies
  • 4 views
  • Added October 23, 2026
ai-agentsgo

Security analysis

A100/100

Scanned October 23, 2026

npx -y skills add jammyfu/open-game-skills --skill puzzle-design --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Puzzle Design?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Puzzle Design
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/jammyfu-puzzle-design/badge)](https://www.skillsdirectory.com/skills/jammyfu-puzzle-design)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: puzzle-design
description: Use when a puzzle needs explicit legal verbs, properties, solution policy, reset behavior and evidence without assuming one room, one answer or one teaching sequence.
---

# Puzzle design

Ask the intended structure: authored-one, multi-solution, systemic-toy, sequence, hybrid, or existing.

## Contract

Publish:
- stable `puzzle_id` and revision
- legal verb/property set and their owners
- legal solution policy
- state variables and stable state IDs where needed
- fail/reset/leave policy
- hint policy and accessibility alternatives
- evidence for intended and allowed alternate solutions

## Rules

1. A puzzle may teach, test, combine or remix properties; the number of rooms or new combinations is project-authored.
2. An allowed improvised solution remains valid when it satisfies the published property/state rules.
3. Failure/reset behavior must avoid unrecoverable state unless that is an explicit project rule.
4. Hinting may range from none to explicit solution guidance; evaluate it against the project's learning/accessibility goal rather than declaring every direct hint a design failure.
5. Chemistry, physics, inventory, gates and narrative state stay owned by their respective skills.

## Acceptance

Given a `puzzle_id`, revision and start state, a reviewer can identify legal verbs/properties, accepted solution policy, reset path, hint policy and which outcomes are valid. Different presentation or room layout does not silently change the puzzle's rules.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

View all in ai-agents →

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.