dsh-office-artifacts
Skill4 starsChanged 3 months agoMITView on GitHub
Create, repair, transform, or verify XLSX, DOCX, PPTX, and PDF deliverables with format-aware tools and reopen or render checks.
In this repository18 other files it gives its agents.
- dsh-changelogSkill
- dsh-chinese-docsSkill
- dsh-code-reviewSkill
- dsh-debug-sessionSkill
- dsh-dependency-auditSkill
- dsh-doc-syncSkill
- dsh-git-commitSkill
- dsh-plugin-clientSkill
- dsh-plugin-devSkill
- dsh-plugin-i18nSkill
- dsh-plugin-publishSkill
- dsh-preset-authoringSkill
- dsh-pr-reviewSkill
- dsh-refactor-safeSkill
- dsh-skill-adapterSkill
- dsh-task-breakdownSkill
- dsh-test-firstSkill
- dsh-webapp-testingSkill
The file
Files
DSH Office Artifacts
- Identify the output format, reader, sources, template, required content, style constraints, and recipient application.
- Preserve source files. Never overwrite an original unless the user explicitly requests it; write a new output path by default.
- Inspect installed tools and use a format-aware library or application. Do not edit OOXML archives through raw string replacement except for a tested, narrow repair.
- Structure content before layout and retain sources for claims, calculations, and externally supplied data.
- Reopen every saved artifact with a parser or native-capable library. Render or visually inspect layout-sensitive output.
- For XLSX, preserve formulas as formulas, keep dates and numbers typed, and check formulas, sheet names, filters, frozen panes, and chart references.
- For DOCX, use styles and preserve heading hierarchy, links, lists, table semantics, citations, and page-break intent.
- For PPTX, preserve the template theme and layouts, then check text overflow, contrast, image crops, charts, and reading order after rendering.
- For PDF, confirm page count, order, extracted text or OCR, links, form values, and redactions. A visual black rectangle is not a verified redaction.
- Report output path, source inputs, checks performed, known limitations, and any required human review.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

