← Back to Skills Marketplace
Word Comments Extractor
by
Tomyan-Zhang
· GitHub ↗
· v1.0.2
· MIT-0
100
Downloads
1
Stars
0
Active Installs
3
Versions
Install in OpenClaw
/install word-comments-extractor
Description
Extract comments from Word documents and format them into standardized review opinions. Auto-matches page numbers, agent-powered semantic polishing. Designed...
Usage Guidance
This skill appears internally consistent and runs entirely on your machine, but before installing/using it consider: 1) Run it on Windows with Microsoft Word installed and install pywin32 as instructed. 2) Opening documents via the Word COM interface can cause Word to resolve external links or, in some document types, trigger macros — avoid running on untrusted documents or run inside a sandbox/VM. 3) The skill does not declare any network calls or credentials, but the agent will see the extracted comment text during polishing; if comments contain sensitive data, treat outputs and conversation history accordingly. 4) The package source/homepage is not authoritative in the metadata provided; if you require provenance, review the included script (extract_comments.py) yourself before running or request a skill with a verified upstream repository.
Capability Analysis
Type: OpenClaw Skill
Name: word-comments-extractor
Version: 1.0.2
The skill is a legitimate tool for extracting and formatting comments from Microsoft Word documents. The Python script (extract_comments.py) uses standard libraries and the Word COM interface (pywin32) to retrieve comment text and page numbers locally, with no evidence of network activity, data exfiltration, or malicious execution. The instructions in SKILL.md are focused entirely on semantic text polishing and do not contain any prompt-injection attacks or harmful directives.
Capability Assessment
Purpose & Capability
Name/description, SKILL.md, and the Python script all align: the script unpacks .docx, extracts comments and anchor text, and uses Word COM (pywin32) to obtain page numbers. Requiring Microsoft Word and pywin32 is coherent with the stated page-number matching capability.
Instruction Scope
Instructions are narrowly scoped: run the included script on a local .docx and then let the agent polish the resulting JSON. There are no references to external endpoints or unrelated files/credentials. Note: the script opens the document through Word COM, which can have side effects inherent to launching Word (e.g., Word may resolve external links or run embedded macros in certain formats). This is an operational risk of using Word automation rather than an inconsistency with the skill's purpose.
Install Mechanism
No install spec beyond an instruction to pip-install pywin32 (a common Windows package for COM access). There are no downloads from arbitrary URLs or archive extraction beyond the .docx unpacking done locally by the script.
Credentials
The skill requests no environment variables, no credentials, and no config paths. That is appropriate for a local document-processing tool that does not interact with cloud services.
Persistence & Privilege
The skill is not forced-always, does not request elevated platform-wide persistence, and does not modify other skills' configs. It runs only when invoked.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install word-comments-extractor - After installation, invoke the skill by name or use
/word-comments-extractor - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.2
word-comments-extractor 1.0.2
- Added explicit platform ("win32") and binary dependency ("Microsoft Word") metadata.
- Added an install section detailing the required `pywin32` installation and manual Microsoft Word requirement.
- No changes to core functionality or usage instructions.
v1.0.1
- English-language documentation and output examples added; all content now in English.
- Improved clarity on architecture: Python script now unpacks `.docx` and extracts page numbers in a single step; agent performs all rewriting.
- Usage simplified: extraction script takes only the `.docx` path; less manual XML handling required.
- Output format standardized for English: clearer, numbered suggestions per page, matching typical investment banking/legal expectations.
- Original intent, quality standards, and prohibited rewriting patterns remain enforced in new documentation.
v1.0.0
Initial release with core comment extraction and polishing functionality:
Automatically extracts comments and anchor text from Word documents with accurate physical page number matching
AI-powered semantic understanding, professional polishing, and formatted output of comment content
Supports investment banking QC standards: auto-generated descriptions, numbering, and standardized professional phrasing
Output checklist designed for review and audit, adaptable to investment banking, legal, and other document review scenarios
初始版本发布，实现核心批注提取与润色功能：
- 从Word文档中自动提取批注及锚定文本，准确匹配物理页码
- 由AI进行批注内容的语义理解、专业润色与格式化输出
- 支持投行质控标准：自动提炼描述、编号、统一专业措辞
- 输出清单便于复核、审核，适配投行、法律等多种场景
Metadata
Frequently Asked Questions
What is Word Comments Extractor?
Extract comments from Word documents and format them into standardized review opinions. Auto-matches page numbers, agent-powered semantic polishing. Designed... It is an AI Agent Skill for Claude Code / OpenClaw, with 100 downloads so far.
How do I install Word Comments Extractor?
Run "/install word-comments-extractor" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Word Comments Extractor free?
Yes, Word Comments Extractor is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Word Comments Extractor support?
Word Comments Extractor is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Word Comments Extractor?
It is built and maintained by Tomyan-Zhang (@tomyan-zhang); the current version is v1.0.2.
More Skills