Tesseract OCR文字识别

基于Tesseract引擎的OCR文字识别技能，支持中文、英文、中英混合三种模式，输出text/structured/question_answer三种格式。

This skill appears to be a local Tesseract OCR tool and does not contact external servers or ask for credentials, but there are important practical issues to consider before installing: - Install Tesseract on the host first (the SKILL.md mentions it, but the skill's metadata did not declare this dependency). Without the Tesseract binary the skill will fail. - The included Python has bugs: a duplicated function, a missing import (os) and a relative import that may fail when run as a script. Expect runtime errors unless the code is fixed. - Because the package metadata is inconsistent (version mismatch, differing descriptions), treat this as low-quality packaging rather than a polished official skill. - Run and review the code in an isolated/sandboxed environment before enabling autonomous invocation. Test the CLI on non-sensitive images to confirm behavior. If you need a stable OCR skill, prefer one that explicitly declares system dependencies (Tesseract binary) in metadata, has matching versions, and has been tested to run without code errors.

Type: OpenClaw Skill Name: tesseract-ocr-skill Version: 1.0.1 The skill provides legitimate OCR functionality using the Tesseract engine, specifically tailored for extracting structured text and Q&A pairs from images (e.g., TCM textbooks). The code in `scripts/tesseract_ocr.py` and `tesseract_ocr_skill.py` uses standard libraries like pytesseract and Pillow without any evidence of malicious intent, data exfiltration, or prompt injection. A minor bug exists in `scripts/tesseract_ocr.py` where a duplicate function definition uses the `os` module without importing it.