heavy-file-ingestion

What problem does it solve?

Agents and workflows waste model context and cost when they read large or structured files raw. This skill ensures heavyweight files are converted into compact, deterministic artifacts and a lightweight index before any model reasoning occurs.

Core Features & Use Cases

• Convert PDFs, DOCX, PPTX, XLSX, CSV, and TSV into markdown, per-sheet CSVs, and a small index that surfaces structure and quality flags.
• Create index.md and index.json with previews, stats, quality warnings, and a recommended next step so the main agent can decide whether to read the artifact or escalate.
• Use Case: Normalize a multi-sheet financial workbook into CSVs and a manifest for downstream analysis, or extract a long slide deck into a slide outline and speaker notes for summarization.

Quick Start

Run the bundled Python converter on the target file and then read the generated index.md to decide whether to read the extracted artifacts or escalate to a stronger converter.