Precision PDF MCP Server
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Precision PDF MCP Serverinspect the answer slots in /home/user/worksheet.pdf"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
๐ฏ Precision PDF MCP Server (precision-pdf-mcp)
Pixel-Perfect PDF Form and Worksheet Filling for AI Coding Agents.
Powered by Hybrid Vector-Raster Geometry, Set-of-Marks (SoM), and Virtual Excel Grounding.
๐ก Why Precision PDF?
Filling assignment worksheets, tax forms, and scanned PDF workbooks has always been a painful failure mode for Large Language Models (LLMs):
โ The "Floating Text" Bug: LLMs guess raw
(x, y)floats, causing answers to float aimlessly above or below lines.โ The "Collision" Bug: Answers smash directly into printed text (e.g.,
Heis going,Parentsmust,1)wear).โ The "Scanned PDF" Blindspot: Most PDF tools rely only on vector drawings. Scanned workbooks have 0 vector lines, leaving AI completely blind.
โ Multi-Line Truncation: Long answers get crammed into line 1 or truncated because the agent doesn't realize two lines are printed for that question.
Precision PDF solves this once and for all.
Related MCP server: ReadPDFx - OCR PDF MCP Server
๐ Key Innovations
[ Input PDF: Digital or Scanned Workbook ]
โ
โผ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ 1. Tri-Layer Hybrid Detector โ
โ โข Vector Lines & Rects (pdfplumber) โ
โ โข In-Memory Morphological Line Scanner (NumPy) โ
โ Detects physical lines in scanned images in RAM! โ
โ โข Text Anchors & Gap Numbers (1), 2., He_________) โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโฌโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ
โผ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ 2. Context Guards & Smart Layout โ
โ โข Padding Guard: Auto +6pt..+10pt margin after โ
โ printed subjects (He, You, Jenny...) โ
โ โข Multi-Line Grouping: Detects stacked lines & auto- โ
โ wraps long sentences at natural word boundaries โ
โ โข Auto-Scale Font: Scales down font to prevent crash โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโฌโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ
โผ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ 3. Virtual Excel Matrix & Set-of-Marks (Agent View) โ
โ โข Renders Top Columns (A, B, C...) & Rows (1, 2, 3) โ
โ โข Badges targets: [S01], [S02], [S03]... โ
โ โข STRICTLY VIRTUAL: Only for the agent to inspect! โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโฌโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ
โผ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ 4. Clean Vector Injection (Output PDF) โ
โ โข Baseline Snapped: Text rests 1.2pt above line โ
โ โข Centered in Boxes: Dead-center horizontally & vert โ
โ โข Crisp Circles / Underlines: Vector option markers โ
โ โข ZERO GRID POLLUTION: Output PDF is 100% clean! โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ๐ฆ Installation & Quickstart
Method 1: Instant Run with uvx (Recommended)
No installation required. Run directly in Claude Desktop or Cursor via uvx:
uvx precision-pdf-mcpMethod 2: Install via pip
pip install precision-pdf-mcpMethod 3: From Source (Local Development)
git clone https://github.com/zuan412/precision-pdf-mcp.git
cd precision-pdf-mcp
pip install -e .๐ ๏ธ MCP Client Configuration
1. Claude Desktop
Add to your claude_desktop_config.json:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"precision-pdf": {
"command": "uvx",
"args": ["precision-pdf-mcp"]
}
}
}2. Cursor IDE
Add to Cursor Settings -> MCP Servers (~/.cursor/mcp.json):
{
"mcpServers": {
"precision-pdf": {
"command": "uvx",
"args": ["precision-pdf-mcp"]
}
}
}3. Antigravity / Gemini CLI / Windsurf / Cline
In your mcp_config.json:
{
"mcpServers": {
"precision-pdf": {
"command": "python",
"args": ["-m", "precision_pdf"]
}
}
}๐งฐ Available MCP Tools
hpe_inspect_virtual_grid
Generates a temporary Virtual Excel Matrix inspector image with Set-of-Marks badges ([S01], [S02]...) and returns a JSON catalog of detected slots.
Arguments:
pdf_path(string, required): Absolute path to the PDF.page_number(int, required): 1-based page number.dpi(int, optional, default: 150): Image resolution.
Returns:
inspector_image_path: Path to the image for agent visual inspection.slots: Dictionary of slots containing grid address (e.g.Col D, Row 14), type, baseline, and width.
hpe_fill_slots
Injects answers directly into the target PDF using Slot IDs.
Arguments:
pdf_path(string, required): Absolute path to the PDF.page_number(int, required): 1-based page number.slot_answers_json(string, required): JSON string of answers.Slot map:
{"S01": "must have left", "S02": "could do"}Multi-line wrap:
{"S03": ["first line text", "second line text"]}Option circle:
{"circle_option": "A"}Option underline:
{"underline_option": "Shall"}
output_path(string, optional): Output PDF path. If omitted, safely updates in-place.
Guarantees:
100% baseline snap (rests naturally on the printed line).
Zero grid pollution (no rulers or badges in final output).
hpe_render_clean_verify
Renders the final, clean PDF page to an image for visual QA verification.
Arguments:
pdf_path(string, required): Path to the completed PDF.page_number(int, required): 1-based page number.dpi(int, optional, default: 150): Output DPI.
๐งช Testing
Run the included unit test suite:
python -m unittest discover -s tests๐ก๏ธ Zero Disk Pollution & Safety Guarantee
Precision PDF is engineered to run in memory:
All morphological line detection and raster scanning execute in RAM via NumPy arrays and
io.BytesIO.Temporary inspector previews use OS-standard
tempfile.gettempdir()/precision_pdfor the$PRECISION_PDF_TEMP_DIRenvironment variable.Stdio JSON-RPC communication is 100% isolated: all logs, warnings, and internal debug outputs route strictly to
sys.stderr.
๐ License
MIT License. See LICENSE for details. Built with โค๏ธ for the AI agent pair-programming community.
This server cannot be deployed
Maintenance
Related MCP Connectors
Detect, review and fill existing PDF forms from JSON or Excel; reuse saved templates
Deterministic PDF tools for AI agents: inspect fields, fill forms, merge PDFs, invoices.
Verified OCR with per-value coordinates, plus a workspace agents can file documents into and query.
Turn any PDF form into a fillable one, fill it from data or documents, verify it, and fax it.
Related MCP Servers
- AlicenseNot gradedqualityCmaintenanceEnables AI-powered extraction and analysis of PDF documents with 40+ specialized tools for text, tables, images, layout analysis, security assessment, and document intelligence. Supports both text-based and scanned PDFs with OCR capabilities.308 PyPI10MIT
- AlicenseNot gradedqualityDmaintenanceProvides intelligent OCR and PDF processing capabilities that automatically detect whether PDFs contain digital text or scanned images and apply appropriate extraction methods. Supports text extraction, OCR processing, structure analysis, and batch operations.MIT
- FlicenseNot gradedqualityDmaintenanceEnables deterministic visual and structural analysis of PDF and DOCX documents, extracting measurable evidence such as blur, OCR confidence, and image anomalies for auditable forensic workflows.1-
- FlicenseNot gradedqualityCmaintenanceEnables AI agents to extract structured data from PDFs with confidence scores and provenance, and to search, review, and correct documents via MCP tools, resources, and prompts.-