A three-mode image generation skill that adapts to your environment instead of breaking when the API key is missing. Mode A runs locally with OpenAI-compatible endpoints and saves everything to disk. Mode BDetectS when you're in ChatGPT or another host with native image tools and becomes a prompt engineering layer that hands off perfectly structured prompts to whatever's available. Mode C drops to pure advisor mode when there's no image capability at all, saving high-quality prompts you can copy elsewhere. Ships with 80+ structured templates across 18 categories covering everything from live commerce UIs to exploded product views to travel maps. The check-mode script runs first and tells you which path you're on.

skills.sh installs
6.2K
GitHub stars
14.8K
last updated
Aug 21, 2026

Install

$ npx -y skills add conardli/garden-skills --skill gpt-image-2 --agent claude-code

Installs into .claude/skills of the current project.

Files

GPT Image 2

这是一个面向 GPT Image 2 的聚焦型技能,在 3 种运行环境下都能用,但行为差异显著。第一步必须先确定当前运行模式。

它只做两类图像任务:

  • 生成图片:POST /images/generations
  • 编辑图片:POST /images/edits

本文件保留:运行模式、技能结构、环境变量、保存 / 命名规则、模板索引、模式感知工作流。详细模板全部放在 references/,分层组织:

  • 一级:分类目录
  • 二级:单模板 Markdown 文件

运行模式(必读,做任何事之前先确定)

本 Skill 自带一个轻量探测脚本,先跑一次,再根据结果决定怎么干活:

node skills/gpt-image-2/scripts/check-mode.js
# 想拿结构化结果给上层程序用:
node skills/gpt-image-2/scripts/check-mode.js --json

输出会给出 mode = A / A? / B-or-C 以及 recommendation。三个模式定义如下:

Mode A · Garden 本地生图

触发条件:环境变量 ENABLE_GARDEN_IMAGEGEN 为真(1 / true / yes / on)且 存在 OPENAI_API_KEY。

行为:完整端到端跑通"选模板 → 写 prompt → 调用脚本 → 出图落盘"。

  • 用 scripts/generate.js 文本生图、scripts/edit.js 编辑现有图。
  • prompt 默认落盘到 garden-gpt-image-2/prompt/、图片落盘到 garden-gpt-image-2/image/。
  • 这是最强的模式:你是图像工具的"持有者"。

Mode B · Host-Native 委托宿主出图

触发条件:未启用 Garden(ENABLE_GARDEN_IMAGEGEN 未设置 / 为假),但当前宿主 Agent 自带图像生成工具或图像 MCP。

典型识别信号(你应该自检):

  • 你的工具集里出现 image_generation / imagegen / dalle / nano_banana / mcp__*image* / make_image / 类似名字
  • 用户在 ChatGPT / Codex / Gemini / Cursor 等支持原生出图的客户端中调用本 Skill
  • 用户显式说"用你自己的工具出图"

Sponsors

Details

Repository
github.com/conardli/garden-skills
Category
AI & Agent Building
First seen
May 16, 2026

More from conardli/garden-skills

All 5
1assistant-ui avatarupdateassistant-ui/skillsUpgrades an existing assistant-ui project to current releases and executes the resulting migrations. Use when the user wants to update, upgrade, bump, or migrat6Kskills.sh installs
2aws avataramazon-aurora-postgresqlaws/agent-toolkit-for-awsAmazon Aurora PostgreSQL — creates, modifies, and advises on Aurora PostgreSQL clusters specifically (PostgreSQL-compatible engine, Aurora serverless, express c6Kskills.sh installs
3aws avatarquerying-aws-s3aws/agent-toolkit-for-awsQueries S3 object metadata, tracks bucket activity, audits object changes, searches annotations, and analyzes storage metrics using S3 Metadata system tables (j6Kskills.sh installs
4forcedotcom avataragentforce-generateforcedotcom/sf-skillsBuild, modify, optimize, debug, and deploy agents with Agentforce Agent Script. TRIGGER when: user creates, modifies, optimizes, or asks about .agent files or a6Kskills.sh installs
5forcedotcom avataragentforce-architecture-analyzeforcedotcom/sf-skillsDeclared architecture snapshot for one Agentforce agent: planner, topics, actions, flows, Apex, prompt templates, and NGA plugins. Renders a human-readable arch5.9Kskills.sh installs
6forcedotcom avataragentforce-testforcedotcom/sf-skillsWrite, run, and analyze structured test suites for Agentforce agents — functional AND security. TRIGGER when: user writes or modifies test spec YAML (AiEvaluati5.8Kskills.sh installs