How to use the Build Game Inventory agent skill

What does the build-game-inventory skill make the agent do?

The skill is a SKILL.md file of instructions that an AI coding agent reads when your request matches its description. It covers item schemas, pickup flows, stack rules, equipment slots, atomic swaps, save migration, and no-loss regression testing.

The core rules are: keep one typed, serializable item source of truth; reference item definitions by stable ID instead of copying stats or descriptions into UI code; and treat every transfer (pickup, equip, unequip, swap, drag and drop, consume, sell) as a transaction.

• Model stable IDs, definitions, ownership, stack limits, equipment compatibility, effects, rarity, presentation data and migration version as separate parts.
• Validate source ownership and destination legality before removing anything.
• Compute the full next inventory and equipment state, then commit once or reject with no change to either side.
• Persist only after the committed state is valid.
• Expose slot compatibility, equipped state, quantity, tooltip facts and failure feedback, with keyboard and touch alternatives to drag-only actions.
• Write tests for invalid destinations, full inventory, duplicate pickup, mid-action swap, save/load, migrations, reset and repeated dispatch.

Install the skill from the ZIP

Download the ZIP for this skill from this page and unzip it. The package contains agents/openai.yaml and SKILL.md. Make sure the folder ends up named build-game-inventory so it contains SKILL.md at its top level.

For a personal install, place the folder in ~/.claude/skills/. To limit it to one game project, place it in that project's .claude/skills/ folder instead.

If you use the Claude apps, you can instead upload the skill under Settings → Capabilities → Skills.

How do you trigger the skill with a natural request?

You do not need a special command. The agent loads the skill when your request fits its description, so describe the inventory problem in plain language. Mention the game, the engine or language if you know it, and the specific symptom.

Requests that match the skill's scope include asking for an item schema, a pickup flow, an equipment swap that does not lose items, drag-and-drop inventory UI, or a save migration for existing items.

• "Add an equipment system with head, chest and weapon slots. Swapping gear must never lose or duplicate an item."
• "Players are duplicating potions when they swap items mid-combat. Find the cause and fix it."
• "Write a save migration that moves our old flat item list into the new typed item definitions."
• "Add regression tests that check total item count is conserved across save and load."

What kind of results should you expect?

The skill steers the agent toward a specific design: definitions referenced by ID, transfers that either fully commit or fully reject, and tests that verify item identity is conserved except for explicit consumption or reward rules. Expect the agent to explain which transfers it made atomic and which tests it added.

The skill gives guidance and patterns. It does not provide a packaged engine module, so the agent still writes code that fits your game's architecture, and you should review it and run your own tests.

Safety notes: scripts and network

According to the listing, this skill does not run scripts and does not use the network. It requires nothing beyond the agent itself. The files are agents/openai.yaml and SKILL.md.

Because it is only instructions, the risk is mainly in the code the agent writes. Review changes to transaction logic and save migrations carefully, since a bug there can corrupt player saves. Run the regression tests the skill proposes before shipping.

Tips for getting better results

Give the agent your current item data shape and save format. Migration work goes much better when the old and new schemas are both visible.

Name the transfer that is failing (for example, equip swap or sell) so the agent applies the transaction pattern to that path first.

Ask for tests before refactoring. Regression tests written against the current behaviour show whether a change loses items.

• Use the skill for item and inventory logic, not for unrelated game systems.
• Keep your own engine or UI framework conventions in the request so the output fits your codebase.