中文文档 | Local-first PDF processing for AI agents.
pdf-toolbox-mcp is an MCP project maintained in the twoer/pdf-toolbox-mcp repository. The maintainer describes its purpose as “中文文档 | Local-first PDF processing for AI agents”. Read the project documentation for the complete scope.
The documented pypi package is pdf-toolbox-mcp. The configuration below comes from the maintainer’s example; adapt its paths and account values to your environment.
Project documentation · Repository · Discovery listing. This profile records the public documentation checked on 2026-10-08.
Check the linked publisher documentation and your connected client's tool list. Names and availability can vary with the server version and account permissions.
| Tool | What it does |
|---|---|
| Single | file outputs are written to a same-directory temporary file and atomically replaced only after the operation succeeds. |
| pdf_info | Pages, encryption status, metadata — always call first |
| is_searchable | Smart routing: text density check → recommends extract_text or ocr_pdf |
| extract_text | Layout-aware text, exact page ranges 1-3,5, per-page mode |
| ocr_pdf | OCR write-back: scan → searchable PDF (deskew, skip/redo, lang fallback) |
| batch_ocr | Whole-directory OCR with per-file results, retries, timeouts |
| render_pages | PNG per page, return_images=true streams image blocks to the vision model |
| extract_images | Pull embedded images (inventory or PNG files) |
| extract_attachments | Pull embedded attachment files |
| list_fonts | Font audit — non-embedded fonts risk missing glyphs on other machines |
| unlock_pdf | Decrypt with user password, output a clean decrypted file; password sent via stdin |
| protect_pdf | AES-256 + granular permissions (print/extract/modify/…) |
| split_pdf | By ranges or every N pages |
| merge_pdfs | Ordered merge |
| rotate_pages | 90/180/270 on selected pages |
| check_repair | Structural check; repair=true rebuilds damaged files |
| linearize | Web-optimized progressive-loading output |
| sanitize | Publishing hygiene: strip JS/OpenAction/metadata/attachments |
| redact | True redaction: affected pages rasterized + opaque boxes — redacted text physically unrecoverable, other pages keep their text layer (rasterize_all=true for… |
| redact_text | Redact by content: locate every occurrence of the given keywords and black them out — no manual coordinates needed |
| locate_text | Find where text occurs: page + bounding boxes (PDF points, top-left origin) — the foundation for redaction & highlighting |
| fill_form | Fill AcroForm fields (missing fields reported) |
| edit_metadata | Set/clear Title/Author/… (docinfo + XMP) |
| compress_pdf | Compress, optionally down a quality ladder until hitting target_mb |
| dependency_status | Probe system tools + install commands |
| doctor | One-shot onboarding check: imports, dependency probe, README paths |
{
"mcpServers": {
"pdf-toolbox": {
"command": "uvx",
"args": [
"--from",
"pdf-toolbox-mcp",
"pdftoolbox"
]
}
}
}From the project documentation. Replace placeholders and example paths before connecting.
Generates both vector and raster images using EverArt's AI models, providing flexible storage and format conversion.
Provides file system operations, analysis, and manipulation through a standardized tool interface.
1AIVault is a local-first desktop app that acts as a portable memory vault for AI tools. It stores your memories, decisions, preferences, and reusable skills in a local SQLite database and exposes…
A Model Context Protocol server
Expose the 5dive agent-fleet CLI as Model Context Protocol tools. Point any MCP client (Claude Desktop, Cursor, Cline, or your own) at this stdio server to file tasks, inspect and message agents, and read the fleet…
Automated DOCX Redlining Engine