Product exports and AI-agent integration

PDFEazy accepts portable files from your existing tools. It does not sign into product accounts or fetch private data automatically. Common file formats provide compatibility; they are not a claim that every product API has a dedicated live adapter.

Supported files

• PDF: combine selected pages, rotate, index and number. Review restrictions for signed and annotated files.
• JPG and PNG: images are fitted to A4 pages. No OCR or HEIC support in this release.
• JSON, CSV, Markdown and plain text: readable report pages, up to 250 KB per file, with the complete original file attached.

LiquiLens product family

Use PDF or supported data exports from LiquiLens, Seiche, Undertow, Palimpsest, Financial Evidence, Evidence Carrier, Wake and NarcoScope. The text-export path prints supplied fields without inventing missing values or changing source statuses. It preserves original UTF-8 bytes in an attachment; the page font displays basic Latin characters only.

For Financial Evidence, save a JSON QueryResult from your configured interface. Fields such as as_of, source_url, rights_status, availability and carrier_verification remain in the export. A new PDF does not refresh those timestamps or grant permission to redistribute the data.

Spreadsheet workflows

Use XLSX or UTF-8 CSV/TSV exports with Excel & spreadsheets. Clean, combine, compare or group data locally, then export XLSX, CSV, typed JSON or an eligible PDF table. Optional change records stay separate. The new local source build also registers spreadsheet_inspect and spreadsheet_process; they are not yet in the published npm release. See the AI handoff guide and provenance guide.

AI agents and local automation

The local PDF Suite MCP server provides document operations without a document-upload service. This website does not expose a public path-based PDF API.

{
  "mcpServers": {
    "pdf-suite": {
      "command": "node",
      "args": ["/absolute/path/to/pdf-suite-mcp/dist/index.js"]
    }
  }
}

After building this source checkout, agents can call pdf_build_packet with a list of local sources and a new output directory. The previously published npm version does not yet contain this new tool.

{
  "sources": [{"filePath": "/reports/review.json"},
              {"filePath": "/reports/source.pdf", "pages": [1, 3]}],
  "outputDir": "/reports/output/new-review",
  "title": "Evidence review",
  "includeIndex": true,
  "numberPages": true,
  "prefix": "REVIEW-"
}

The result includes packet.pdf and file-record.json. Existing output directories are rejected to protect previous work. Future service integrations should use this versioned file contract rather than screen scraping.