$npx -y skills add dkyazzentwatwa/chatgpt-skills --skill document-converter-suiteConvert PDFs, Office docs, markdown, HTML, and tables between editable formats. Use for best-effort document conversion, PDF utilities, form filling, and table extraction.
| 1 | # Document Converter Suite |
| 2 | |
| 3 | Run best-effort extraction and rebuild workflows across common document formats. Preserve clean structure, not pixel-perfect layout. |
| 4 | |
| 5 | ## Use This For |
| 6 | |
| 7 | - Converting between `pdf`, `docx`, `pptx`, `xlsx`, `txt`, `csv`, `md`, and `html` |
| 8 | - Pulling tables or spreadsheet-style grids into editable outputs |
| 9 | - Running utility PDF operations such as merge, split, rotate, watermark, or page extraction |
| 10 | - Filling simple document or form-style templates |
| 11 | |
| 12 | ## Workflow |
| 13 | |
| 14 | 1. Confirm the source format, target format, and whether editability or fidelity matters more. |
| 15 | 2. Use `scripts/convert.py` for single documents and `scripts/batch_convert.py` for folders. |
| 16 | 3. Use the bundled utility scripts when the user needs a focused PDF or table task: |
| 17 | - `scripts/pdf_toolkit.py` |
| 18 | - `scripts/table_extractor.py` |
| 19 | - `scripts/form_filler.py` |
| 20 | 4. Say explicitly when the output is best-effort and likely to lose layout, images, OCR text, or advanced formatting. |
| 21 | |
| 22 | ## Guardrails |
| 23 | |
| 24 | - Do not promise visual fidelity. |
| 25 | - Treat scanned PDFs as OCR problems, not conversion problems. |
| 26 | - Raise safety caps gradually on large sheets or documents instead of processing everything blindly. |
| 27 | |
| 28 | ## References |
| 29 | |
| 30 | - `references/conversion_matrix.md` for supported paths. |
| 31 | - `references/limitations.md` for failure modes and tradeoffs. |