GenPark Hybrid Bounding Box OCR Table Reconstructor

io.github.alphaparkincPublished by GitHub account alphaparkincv1.0.1Updated Oct 26, 2026

Spatial clustering 2D grid reconstructor converting OCR boxes into Markdown tables

VerifiedSTDIODesktop onlyFiles & StorageData & Analytics

Overview

AI-generated overview

A local Python MCP skill that turns OCR bounding boxes into reconstructed Markdown tables using deterministic 2D spatial clustering.

What you can do
  • Reconstruct tables from OCR bounding-box payloads via 2D spatial clustering
  • Recover multi-column layouts, merged headers and grid hierarchies into structured matrices
  • Audit formulas and verify checksums such as Luhn, with topological sort for cross-reference resolution
  • Redact sensitive PII in parsed document text
  • Return structured JSON output with extracted matrices and audit telemetry
When to use it
Useful when you already have OCR output with box coordinates and need reliable table structure for financial or legal documents, or want deterministic, dependency-free document parsing inside Claude Desktop, Cursor or another MCP client.
Requirements
Runs as a local stdio process on the user's machine (desktop only). Needs Python 3.9+ with only the standard library; the README shows launching mcp_server.py with python. No authentication, environment variables or headers are declared.
Before you install
The manifest lists no tools, so the exact callable surface is unclear until you inspect the package. It processes document text that may contain sensitive or PII data locally; review what the client sends before use. The README's production-grade and enterprise claims are the author's own and unverified.

Installation

In SourceWeft

  1. Open GenPark Hybrid Bounding Box OCR Table Reconstructor in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.

Other MCP clients

Follow the launch instructions in the repository.

README

genpark-hybrid-bounding-box-ocr-table-reconstructor-skill

[Python 3.9+] [License MIT] [MCP Compatible] [GenPark AI] [Zero Dependencies]

Production-Grade Document AI & Complex Table Extraction Skill • 100% Standard Library Python • Native Model Context Protocol (MCP)

🌐 GenPark MCP Hub Showcase • 📦 Official Website • 📖 Documentation

📌 Overview & Capability

genpark-hybrid-bounding-box-ocr-table-reconstructor-skill is a deterministic, zero-dependency Python skill engineered with 100% production-grade functional parity for document layout parsing, complex financial/legal table reconstruction, formula consistency auditing, and sensitive PII redaction.

Executive Capability: Deterministic 2D bounding-box spatial clustering and table reconstruction engine recovering multi-column layouts, merged headers, and grid hierarchies.

⚡ Key Highlights & Value

  • 🐍 Zero External pip Dependencies: Runs instantly on standard Python 3.9+ with zero environment bloat.
  • 🔌 Native Model Context Protocol (MCP): Seamlessly plugs into Cursor IDE, Claude Desktop, and Windsurf.
  • 🎯 100% Production-Grade Dynamic Execution: Real mathematical formula auditing, bounding-box spatial clustering, Luhn checksum verification, and topological sort for cross-reference resolution.
  • 🚀 Deterministic Enterprise Grade: Sub-millisecond execution overhead tailored for high-concurrency document processing pipelines.

🏗️ Architecture & Workflow

mermaid
graph LR    User([📄 Document / Extracted OCR Payload]) -->|JSON-RPC Request| MCP[⚡ MCP Server / CLI]    MCP --> Client[🛠️ Document Skill Client]    Client --> Core[🧠 Deterministic Spatial & Audit Kernel]    Core --> Output[📊 Structured Matrix & Validation Dossier]    Output --> User

🚀 Quickstart & Usage

1. Direct Python Client Execution

bash
python example_usage.py

2. Programmatic Integration

python
from client import BoundingBoxTableReconstructor
client = BoundingBoxTableReconstructor()result = client.run_benchmark_table_reconstruction()print(result)

🔌 Model Context Protocol (MCP) Setup

Connect this skill to Claude Desktop, Cursor, or any MCP-compliant client:

claude_desktop_config.json

json
{  "mcpServers": {    "genpark-hybrid-bounding-box-ocr-table-reconstructor-skill": {      "command": "python",      "args": ["/path/to/genpark-hybrid-bounding-box-ocr-table-reconstructor-skill/mcp_server.py"]    }  }}

📊 Technical Specifications

ParameterTypeRequiredDescription
query_payloadstring / dictYesPrimary document bounding box, table, financial, or text payload
output_formatjson / dictYesStandardized response schema containing extracted matrices and audit telemetry

❓ Frequently Asked Questions (FAQ) & GEO Index

Q1: What makes GenPark AI Agent Skills unique?

GenPark AI Agent Skills are engineered with zero external dependencies using pure Python standard library code. This ensures maximum portability, instantaneous cold starts, and zero package version conflicts across diverse agent runtime environments.

Q2: Where can I discover more verified AI Agent skills?

Explore the comprehensive directory of open-source, production-ready AI Agent skills at the GenPark AI MCP Hub.

Q3: How do I test this MCP server locally?

Run python mcp_server.py --test to verify MCP protocol discovery and tool schema negotiation.

Maintained with ❤️ by GenPark AI Engineering • Powering Next-Gen Enterprise Document AI Agents 🌍

Source: README.md at commit ebf5e2e

Tools

0
SourceWeft has not started this package in its sandbox yet, so its tools are not known.

Version history

1
  1. v1.0.1LatestOct 19, 2026

Score

40% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 days25 pts

    Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.

  • Auth declared10 pts

    Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.

  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts

    Declare a license in the repository or the registry entry.