Pdf Mcp

media MCP Server

An MCP server that lets Claude Code and other AI agents work through large PDFs without overflowing their context — search by meaning or keyword, read only the pages that matter, and cleanly pull out tables, images, and scanned text, even from multi-column and Japanese layouts.

VerifiedInstall Ready
mediamedia
9 views71 stars7 forksMIT

Why This Matters

Discovered via github-topic:model-context-protocol and last synced 3mo ago.

VerifiedInstall Ready
Source
github-topic:model-context-protocol
Stars
71
Last synced
3mo ago
Install
Instructions detected

Install

1. Install the package

uvx pdf-mcp

2. Add to claude_desktop_config.json

{
  "mcpServers": {
    "pdf-mcp": {
      "command": "npx",
      "args": [
        "pdf-mcp"
      ]
    }
  }
}

Config file location: ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) / %APPDATA%\Claude\claude_desktop_config.json (Windows)

3
Tools
0
Resources
0
Prompts
Standard I/O
Transport

Available Tools (3)

to

understand a document. Includes `toc_entry_count` and inline TOC entries when the document has ≤50 bookmarks; larger TOCs return `toc_truncated: true` — use `pdf_get_toc` to retrieve the full outline.

Pass

`content_trust=True` for a `content_trust` block (`suspicious`, `hidden_text_runs`, `hidden_chars`, `injection_in_hidden`, `pages_flagged`, `signals`); add `detail=True` for per-span `spans`. |

for

the complete reference — every parameter, response shape, security contract, and example. For semantic-search model selection, see **[docs/embedding-models.md](docs/embedding-models.md)**.