By name: pdf version: 1.0.0 description: > Extract text and metadata from PDF files using the exec environment. classification_ceiling: CONFIDENTIAL requires_tools: - run_command - write_file - read_file network_domains: - registry.npmjs.org
PDF Text Extraction
Extract text from PDFs using the exec environment.
Setup (One-Time)
run_command: { command: "npm install pdf-parse", cwd: "{workspace}" }
Check if node_modules/pdf-parse exists before reinstalling.
Extract Text
- Write extraction script:
write_file: {
path: "{workspace}/extract-pdf.mjs",
content: "import fs from 'fs';\nimport pdfParse from 'pdf-parse';\nconst buf = fs.readFileSync(process.argv[2]);\nconst data = await pdfParse(buf);\nconsole.log(JSON.stringify({ pages: data.numpages, text: data.text, info: data.info }));"
}
- Run on target PDF:
run_command: { command: "node extract-pdf.mjs /path/to/document.pdf", cwd: "{workspace}" }
Specific Pages
Split by form feed (\f) and select ranges for large PDFs.
Key Behaviors
- Check if pdf-parse installed before using
- PDF content inherits session classification — may be sensitive
- For 100+ page PDFs, extract specific ranges
- Garbled output = scanned/image PDF, inform user OCR is needed
- Local processing only — no network calls for extraction