openclaw 网盘下载
OpenClaw

技能详情(站内镜像,无评论)

首页 > 技能库 > MiniMax PDF OCR

使用 MiniMax Vision API 识别 PDF/图片中的文字

媒体与内容

许可证:MIT-0

MIT-0 ·免费使用、修改和重新分发。无需归因。

版本:v1.0.0

统计:⭐ 0 · 41 · 0 current installs · 0 all-time installs

0

安装量(当前) 0

🛡 VirusTotal :可疑 · OpenClaw :可疑

Package:chongjie-ran/minimax-pdf-ocr

安全扫描(ClawHub)

  • VirusTotal :可疑
  • OpenClaw :可疑

OpenClaw 评估

The skill appears to implement PDF→image→OCR via a remote MiniMax Vision API (which matches its stated purpose), but there are multiple metadata/instruction inconsistencies you should understand before installing.

目的

The code and SKILL.md implement PDF→PNG conversion (pdftoppm/poppler) and send images to a MiniMax Vision API for OCR — this aligns with the skill name/description. However, the registry metadata (which claimed no required env vars or binaries) is inconsistent with the SKILL.md and code that require an API key (MINIMAX_API_KEY) and rely on a system binary (pdftoppm).

说明范围

Runtime instructions are focused: convert PDF to images, base64-encode images, and POST them to https://api.minimax.chat/v1/text/chatcompletion_v2 for OCR, then save Markdown. The instructions do send image data (embedded as data URLs) to an external API — expected for an OCR skill but important for privacy. SKILL.md also instructs installing npm packages (openai, pdf2image) that the shipped code does not use; this is inconsistent but not dire…

安装机制

No install spec (instruction-only) lowers risk. The only non-JS install guidance is to install poppler (provides pdftoppm) via brew — a standard system package. There are no remote download/extract steps or obscure URLs in the install path.

证书

The code requires a single credential (MINIMAX_API_KEY) and optionally OUTPUT_DIR — proportional for a remote OCR API. However, the registry metadata incorrectly lists no required env vars; this discrepancy between declared requirements and actual code is a red flag (could be sloppy packaging or mis-declared permissions). No other credentials are requested.

持久

The skill does not request persistent/always-on privileges and does not modify other skills or system-wide configs. It runs as a user-invoked Node script and only accesses the files you provide plus the environment API key.

pdf-ocr-minimax.js:36

Shell command execution detected (child_process).

pdf-ocr-minimax.js:14

Environment variable access combined with network send.

pdf-ocr-minimax.js:58

File read combined with network send (possible exfiltration).

安装(复制给龙虾 AI)

将下方整段复制到龙虾中文库对话中,由龙虾按 SKILL.md 完成安装。

请把本段交给龙虾中文库(龙虾 AI)执行:为本机安装 OpenClaw 技能「MiniMax PDF OCR」。简介:使用 MiniMax Vision API 识别 PDF/图片中的文字。
请 fetch 以下地址读取 SKILL.md 并按文档完成安装:https://raw.githubusercontent.com/openclaw/skills/refs/heads/main/skills/chongjie-ran/minimax-pdf-ocr/SKILL.md
(来源:yingzhi8.cn 技能库)

SKILL.md

打开原始 SKILL.md(GitHub raw)

暂无本地缓存内容,可在后台执行详情同步。