word-chinese-automation
AdvisoryAudited by Static analysis on Apr 30, 2026.
Overview
No suspicious patterns detected.
Findings (0)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Incorrect or overbroad corrections could change wording throughout the generated modified document.
The script can apply replacements across the Word document and save a corrected copy; this is aligned with the proofreading purpose, but users should verify AI-created corrections before using the modified file.
para.text = para.text.replace(original, corrected) ... output_path = str(input_file.parent / f"{input_file.stem}_修改{input_file.suffix}") ... doc.save(output_path)Keep the original document, review the proofreading report and corrections JSON, and compare the generated modified copy before distributing or replacing any official file.
Sensitive text from the Word document may remain in local intermediate files or console logs after the proofreading run.
The splitter stores document text in an intermediate JSON file and prints extracted sentences to the console; this is expected for proofreading but duplicates potentially sensitive document content locally.
"full_text": text, "sentence": sentence ... print(f" {item['sentence']}") ... output_path = str(input_file.parent / f"{input_file.stem}_sentences.json")Use this only on documents appropriate for local processing, and delete generated sentence/correction JSON files when they are no longer needed.
A user may need to install an undeclared Python package manually for the skill to work.
The code expects the python-docx package, while the registry lists no install spec or required binaries; this is a setup/provenance note rather than suspicious behavior.
except ImportError:
print("错误：需要安装 python-docx 库")
print("运行: pip install python-docx")Install python-docx only from a trusted package source and consider pinning a known-good version in local environments.