pdf-mcp

Args:
    filename: Path to a PDF file.

When max_level is not specified and the TOC is very large, the depth is automatically
reduced to keep the response manageable. The response will include an
``auto_trimmed_to_level`` field and a hint when this happens.

Args:
    filename: Path to a PDF file.
    parent: Return only children of the entry whose title contains this string
            (case-insensitive). For example, pass a chapter title to get its sections.
    max_level: Only include entries up to this depth (1=chapters only, 2=sections, etc.).
               When used with parent, levels are relative: 1 means direct children only,
               2 means children and grandchildren, etc.

Args:
    filename: Path to a PDF file.
    start_page: First page number (1-indexed, inclusive).
    end_page: Last page number (1-indexed, inclusive). Defaults to start_page.
    format: Output format. "json" returns structured page data with block/line/span
            detail. "text" returns plain text. "markdown" returns markdown via
            PyMuPDF4LLM. "html" returns HTML.
    include_headers_footers: If False, crops top/bottom margins to exclude
                             headers and footers. Default True.

Args:
    filename: Path to a PDF file.
    page: Page number (1-indexed).
    dpi: Image resolution. Default 150 (good balance of readability and size).
    output: "base64" returns the image inline as MCP image content.
            "file" writes to a temp file and returns the path.

Returns a list of hits with page number and surrounding context.

Args:
    filename: Path to a PDF file.
    query: Text to search for.
    context_chars: Characters of context to include around each hit. Default 100.