PDF (HybridClaw)

What it does

Provides a Node/JS-first PDF toolkit for agents: create new PDFs, extract text, render pages to images, detect and fill form fields, place text by coordinates, and merge or split documents. It prefers bundled scripts and enforces a linear, reproducible workflow.

When to use it

Use this skill whenever a user mentions a .pdf file or asks to read, summarize, extract structured fields (invoices/forms), render pages, or fill forms. It's the primary path for PDF tasks inside the HybridClaw workspace and should be chosen before external CLIs or ad-hoc tools.

What's included

• Scripts: bundled Node scripts (extract_pdf_text.mjs, render_pdf_pages.mjs, create_pdf.mjs, fill_fillable_fields.mjs, etc.) — has_scripts: true
• References: forms.md and reference.md present in the skill directory — has_references: false (reference files included)
• Instructions: clear working rules (workspace-root execution, prefer bundled scripts, /tmp for scratch), default extraction workflow, and troubleshooting boundaries if Node scripts fail.

Compatible agents

Best used with agents that support Node/JS execution and bundled script invocation (e.g., HybridClaw runtime, Codex/Code-style agents that can run workspace node scripts).