← 返回 Skills 市场
shengy983

sudocode-nano-banana2

作者 shengy983 · GitHub ↗ · v1.0.1 · MIT-0
cross-platform ✓ 安全检测通过
165
总下载
0
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install sudocode-nano-banana2
功能描述
Generate or edit images with Sudocode's nano-banana2 image service. Use when the user wants text-to-image or image-to-image generation through the bundled Su...
安全使用建议
This package appears to be a straightforward client for Sudocode image generation. Before installing: 1) Confirm you trust the Sudocode service (sudocode.run / sudocode.us) because your API key will be sent there. 2) Note that the initializer will store the API key in ~/.openclaw/.env; if you prefer not to persist secrets to disk, provide the key via environment variables at runtime instead. 3) The skill writes per-run logs beside your output files — clean up sensitive artifacts if needed. 4) Verify the registry metadata mismatch about required env vars (the package does require SUDOCODE_IMAGE_API_KEY). If you have doubts, run the scripts in an isolated environment or review the two Python files (they are short and readable).
功能分析
Type: OpenClaw Skill Name: sudocode-nano-banana2 Version: 1.0.1 The skill is a standard API client for an image generation service hosted at sudocode.run. It includes a setup script (scripts/init_env.py) to manage API keys in a local .env file and a main script (scripts/sudocode_nano_banana2.py) that sends prompts and images to the remote endpoint. All behaviors, including network requests and file system access for logging and output, are directly aligned with the stated purpose of text-to-image generation.
能力评估
Purpose & Capability
The name/description (Sudocode Nano Banana2 image generation) aligns with the provided scripts and runtime behavior: the code posts prompts and optional input images to a Sudocode-compatible endpoint and saves results locally. The MODEL identifier difference is documented in SKILL.md and the code.
Instruction Scope
Instructions remain within the stated purpose. They instruct initializing/storing an API key, calling the bundled script, and sending results back to the user. Note: the initializer writes credentials to ~/.openclaw/.env and the runtime writes logs under a 'logs' directory next to the output file — this is expected for this skill but worth being aware of.
Install Mechanism
No install spec (instruction-only packaging) and no downloaded binaries. The SKILL.md suggests pip-installing 'requests' if missing; this is proportionate to the script's use of the requests library.
Credentials
The only required secret is SUDOCODE_IMAGE_API_KEY (primary credential) and an optional SUDOCODE_BASE_URL; that matches the code and SKILL.md. Minor inconsistency: the top summary in the provided metadata listed 'Required env vars: none', whereas _meta.json and the code declare SUDOCODE_IMAGE_API_KEY as required — confirm the registry metadata before trusting an automated listing.
Persistence & Privilege
The skill is not always-enabled and does not request elevated privileges. It persists the API key to ~/.openclaw/.env (its own config area) and writes per-run logs/images to local disk; these are normal for a client that needs an API key and produces files.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install sudocode-nano-banana2
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /sudocode-nano-banana2 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.1
- Clarified that the skill is an API client for Sudocode's remote nano-banana2 service, not a local inference model. - Updated the description to reflect use of the Sudocode API client and remote model naming. - Added a section explaining the difference between the skill name (`sudocode-nano-banana2`) and the upstream model identifier (`gemini-3.1-flash-image-preview`). - Noted that generated images and logs are written to disk near the output path, and credentials are stored only via the initializer. - Improved instructions for packaging, delivery, and environment setup for consistency with remote API usage.
v1.0.0
Initial public release of sudocode-nano-banana2. - Generate images from text prompts or edit local images using the Sudocode nano-banana2 model. - Includes setup instructions for environment variables and obtaining an API key from https://sudocode.us. - Provides CLI usage examples; supports image-to-image, text-to-image, and custom output paths. - Details on packaging, distributing, and delivering generated images via the OpenClaw "message" tool. - Covers error handling for common API issues and relay instructions for user feedback.
元数据
Slug sudocode-nano-banana2
版本 1.0.1
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 2
常见问题

sudocode-nano-banana2 是什么?

Generate or edit images with Sudocode's nano-banana2 image service. Use when the user wants text-to-image or image-to-image generation through the bundled Su... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 165 次。

如何安装 sudocode-nano-banana2?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install sudocode-nano-banana2」即可一键安装,无需额外配置。

sudocode-nano-banana2 是免费的吗?

是的,sudocode-nano-banana2 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

sudocode-nano-banana2 支持哪些平台?

sudocode-nano-banana2 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 sudocode-nano-banana2?

由 shengy983(@shengy983)开发并维护,当前版本 v1.0.1。

💬 留言讨论