About office-document
A routing skill that maps office-document tasks to the appropriate official Hermes skill. It distinguishes editable document work from read-only extraction and gives format-specific installation guidance.
Key features
- 01Routes DOCX, XLSX, PPTX, and PDF tasks
- 02Separates editing from read-only extraction
- 03Handles mixed-format requests per artifact
- 04Routes scanned PDFs to OCR skill
- 05Provides matching skills CLI commands
Use cases
- 01Choose a Hermes skill for a document task
- 02Route multiple requested formats separately
- 03Select OCR for an image-only PDF
Install
Copy the command and run it in a terminal.
npx skills add https:/After installing, type/office-documentin Claude Code to confirm it loaded.
Details
From the project READMEOffice document routing
Use this skill as a routing decision, not as a replacement for the document engines. The official Hermes skills are separate skills:
| User task | Hermes skill |
|---|---|
Create, read, edit, template, or review Word .docx | docx |
Create, read, or edit Excel .xlsx / CSV | xlsx |
Create, read, or edit editable PowerPoint .pptx | powerpoint |
| Create, read, merge, fill, secure, or otherwise manipulate PDF | pdf |
| Read-only extraction or summarization | Do not route here; read the file directly |
Visual presentation that does not need an editable .pptx | Starchild slide-creator |
Decision rules
- Determine whether the user needs an editable output or only information. Read-only is not a document-production task. Do not install or invoke a document skill merely to extract or summarize file contents.
- Select exactly one format-specific Hermes skill from the table above.
- For a mixed request, route each artifact to its own format-specific skill;
do not use
powerpointfor a PDF ordocxfor a general presentation. - For scanned or image-only PDFs, use Hermes
ocr-and-documentsbefore or instead ofpdf; a PDF skill cannot invent a missing text layer. - After any edit or creation, follow the selected skill's verification steps. Do not claim format fidelity without read-back and, where applicable, render/visual inspection.
Installing the official Hermes skills
These are the official skills from NousResearch/hermes-agent, not similarly
named community skills.
Skills CLI installation
Install only the one skill required by the current task. Do not install all format-specific skills in advance.
Use the matching command:
- DOCX task：
npx skills add NousResearch/hermes-agent --skill docx - XLSX／CSV task：
npx skills add NousResearch/hermes-agent --skill xlsx - Editable PPTX task：
npx skills add NousResearch/hermes-agent --skill powerpoint - PDF task：
npx skills add NousResearch/hermes-agent --skill pdf - Scanned or image-only PDF：install
ocr-and-documentsonly when OCR is actually needed：npx skills add NousResearch/hermes-agent --skill ocr-and-documents
Do not substitute a community package with a similar name unless the user
explicitly asks for that package. After installation, confirm the selected
skill's SKILL.md is present and use its own scripts and verification commands.
Official source mapping
The upstream source paths are:
https://github.com/NousResearch/hermes-agent/tree/main/skills/productivity/docxhttps://github.com/NousResearch/hermes-agent/tree/main/skills/productivity/xlsxhttps://github.com/NousResearch/hermes-agent/tree/main/skills/productivity/powerpointhttps://github.com/NousResearch/hermes-agent/tree/main/skills/productivity/pdfhttps://github.com/NousResearch/hermes-agent/tree/main/skills/productivity/ocr-and-documents
The Hermes docx skill uses python-docx helpers and package health checks;
xlsx uses openpyxl; powerpoint uses python-pptx; and pdf uses
pypdf, reportlab, and pdfplumber. These implementation details describe
the official skills and should not be replaced with guessed local commands.