Agent Skills › jasonkneen/tiny-world-builder › tinyworld-auto-batching

tinyworld-auto-batching

GitHub

定义 Tiny World Builder 自动批处理逻辑，通过缓存模型建议减少调用频率。涉及前端状态管理、API密钥处理及浏览器端验证测试。

.codex/skills/tinyworld-auto-batching/SKILL.md jasonkneen/tiny-world-builder

Trigger Scenarios

修改 Auto 调色板行为或模型推理逻辑调整建议生成或 API 密钥处理流程进行前端自动化功能测试

Install

npx skills add jasonkneen/tiny-world-builder --skill tinyworld-auto-batching -g -y

More Options

Non-standard path

npx skills add https://github.com/jasonkneen/tiny-world-builder/tree/main/.codex/skills/tinyworld-auto-batching -g -y

Use without installing

npx skills use jasonkneen/tiny-world-builder@tinyworld-auto-batching

指定 Agent (Claude Code)

npx skills add jasonkneen/tiny-world-builder --skill tinyworld-auto-batching -a claude-code -g -y

安装 repo 全部 skill

npx skills add jasonkneen/tiny-world-builder --all -g -y

预览 repo 内 skill

npx skills add jasonkneen/tiny-world-builder --list

SKILL.md

Frontmatter

{
    "name": "tinyworld-auto-batching",
    "description": "Use when changing Tiny World Builder Auto palette behavior, model inference, suggestions, or provider\/model\/API-key logic."
}

Tiny World Auto Batching

Auto should not call the model on every placement.

Current intended model:

Capture the current home board as sparse 8x8 JSON using snapshotCells().
Ask the model for a ranked batch of candidate tile actions matching AUTO_SUGGESTIONS_SCHEMA.
Spend autoSuggestionQueue locally for multiple Auto placements.
Refresh after AUTO_REFRESH_EVERY Auto placements or when the queue runs out.
Auto still places on the clicked cell. The model suggests tile/action definitions, not coordinates.

Implementation guardrails:

Keep provider/model/key state shared with the Generate modal via AI_LS.
Use normalizeAutoAction before applying model output.
Use adaptAutoSuggestionToCell(action, x, z) so same-kind clicks increase local intensity.
If no API key exists, open the Generate modal with the Auto key message.
Do not add a backend or new dependency.

Validation:

Stub generateAutoSuggestions in browser eval and verify two Auto suggestions only call the stub once.
Verify no network call is needed while spending cached suggestions.