← 返回 Skills 市场
Minimax Coding Plan Tool
作者
yongjie666888
· GitHub ↗
· v1.0.4
· MIT-0
100
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install minimax-coding-plan-tool-yongjie
功能描述
MiniMax Token Plan工具 - 支持图片生成、图像理解、语音合成和视频生成。直接调用MiniMax官方API，纯JavaScript实现，无需外部MCP服务器。
安全使用建议
This skill is inconsistent and requires caution. What to consider before installing or using it:
- The documentation asks you to provide an OpenClaw OAuth token (sk-cp-...) and the code will send that token to api.minimaxi.com. Treat that token as highly sensitive; do not supply your platform OAuth token unless you explicitly trust the author and have confirmed that the Token Plan really requires platform tokens.
- The SKILL.md advertises TTS and video features that are not implemented in the included code. Ask the author why those features are missing or whether you were given the correct release.
- The tool will read local image files and convert them to base64 for upload. Do not point it at any files that contain secrets or sensitive data (you could accidentally exfiltrate credentials if you rename them to an image extension).
- Metadata mismatches: the embedded _meta.json ownerId/version differ from the registry metadata — this can indicate repackaging or stale metadata. Verify the skill's author and origin before use.
- If you want to test safely: run the code in a network-restricted sandbox and inspect outbound requests (ensure it only calls the intended api.minimaxi.com endpoints). Prefer creating a limited/minimally-privileged API key (not your main OpenClaw OAuth token) if the service supports it.
- If you cannot confirm provenance or necessity of using an sk-cp token, do not provide your platform OAuth token; ask the publisher to provide an official MiniMax-specific API key flow or a privacy justification.
功能分析
Type: OpenClaw Skill
Name: minimax-coding-plan-tool-yongjie
Version: 1.0.4
The skill bundle is a standard implementation of a MiniMax API wrapper for OpenClaw, providing tools for web search, image generation, and image understanding via the api.minimaxi.com endpoint. While there is a discrepancy between the documentation (SKILL.md) and the code (minimax_coding_plan_tool.js) regarding the number of supported tools, and the metadata contains a future publication date, the code logic is transparent, lacks obfuscation, and contains no evidence of malicious intent or unauthorized data exfiltration.
能力评估
Purpose & Capability
The SKILL.md advertises five features (image gen, image understanding, TTS, video generation, web search) but the shipped code only implements web_search, understand_image, and generate_image. The README/header even claims the tool provides only two tools in places — inconsistent claims about capabilities. Required binary (node) and declared env MINIMAX_API_KEY match the implemented code, but the declared scope in the doc (speech/video) is not implemented in code, which is misleading.
Instruction Scope
The runtime instructions explicitly tell users to supply an OpenClaw OAuth token (sk-cp-...) as the API key and show a full 'openclaw config set' example with a long sk-cp token. The code will send that token in an Authorization header to api.minimaxi.com. The tool reads local image files (converts to base64) and will POST that data to the external API — this is expected for image analysis, but it also enables exfiltration of any local file with a supported image extension (an attacker could rename sensitive files to have .png etc.). The SKILL.md and code do not instruct reading other unrelated system paths, but the explicit guidance to use a platform OAuth token is sensitive and risky.
Install Mechanism
There is no install script or remote download — this is an instruction-only skill with a single JS file. Required runtime is node. No external installers or archives are fetched by the skill itself.
Credentials
The skill requires a single env var MINIMAX_API_KEY (declared primary credential). However the documentation instructs using the user's OpenClaw OAuth token (sk-cp-...), which is a highly privileged platform credential. It's unclear whether a platform OAuth token is necessary or whether a separate MiniMax-specific API key could be used. Requesting that platform token and sending it to an external host is disproportionate unless Token Plan truly requires it. Additionally, the SKILL.md includes a hardcoded example token string, which is poor practice and potentially sensitive.
Persistence & Privilege
The skill does not request 'always: true' and does not appear to modify other skills or system-wide settings. It only requires the API key at runtime and will run as a normal user process.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install minimax-coding-plan-tool-yongjie - 安装完成后，直接呼叫该 Skill 的名称或使用
/minimax-coding-plan-tool-yongjie触发 - 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
v1.0.4
- Added detailed Chinese/English documentation for all supported tools: image generation, web search, image understanding, text-to-speech, and video generation.
- Now includes API usage instructions, input/output schemas, and CLI invocation examples for each feature.
- Emphasizes pure JavaScript implementation with direct MiniMax API calls; no external MCP server or dependencies required.
- Describes configuration via `MINIMAX_API_KEY` using OpenClaw OAuth Token.
- Outlines supported models, daily quotas, available voice options, and usage caveats for asynchronous video generation.
元数据
常见问题
Minimax Coding Plan Tool 是什么？
MiniMax Token Plan工具 - 支持图片生成、图像理解、语音合成和视频生成。直接调用MiniMax官方API，纯JavaScript实现，无需外部MCP服务器。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 100 次。
如何安装 Minimax Coding Plan Tool？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install minimax-coding-plan-tool-yongjie」即可一键安装，无需额外配置。
Minimax Coding Plan Tool 是免费的吗？
是的，Minimax Coding Plan Tool 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Minimax Coding Plan Tool 支持哪些平台？
Minimax Coding Plan Tool 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 Minimax Coding Plan Tool？
由 yongjie666888（@yongjie666888）开发并维护，当前版本 v1.0.4。
推荐 Skills