heavy-file-ingestion

Converting large PDF and spreadsheet files into markdown for AI agents requires deterministic conversion of PDFs, DOCX, PPTX, XLSX, and CSV inputs into compact markdown or normalized CSV artifacts before analysis.

The best way to reduce token usage when analyzing heavy documents is to convert them into smaller, structured artifacts first, which prevents agents from wasting context and cost on raw heavyweight files during extraction.

Document conversion handles scanned PDFs or weak text extraction through quality-aware fallbacks that detect poor results, signal warnings in the index, and recommend whether to retry, switch converters, or escalate the task.

Yes, you need dependencies like markitdown, pdfplumber, openpyxl, python-docx, and python-pptx installed, as the Skill detects missing dependencies and signals warnings to ensure deterministic file conversion proceeds correctly.

Yes, you can generate a lightweight index from multiple spreadsheet files that includes source details, preview lines, stats, warnings, and a recommended next step, allowing the agent to route intelligently before reading converted outputs.