deepread

Convert PDFs and images into structured JSON with confidence flags.

1|Updated Feb 1, 2026
One-click install
npx skills add https://github.com/xukrutdonut/openclaw-skills --skill deepread
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: deepread
Source: https://github.com/xukrutdonut/openclaw-skills/tree/main/deepread-ocr
Command: npx skills add https://github.com/xukrutdonut/openclaw-skills --skill deepread

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

AI agents often struggle to extract accurate, structured data from PDFs and images; this Skill automates that process, turning documents into machine-readable outputs and reducing manual data-entry effort.

Core Features & Use Cases

  • Text Extraction: convert PDFs/images into clean, structured data.
  • Structured Data & Confidence: outputs JSON with field values and per-field confidence (hil_flag) for review.
  • Quality Flags & Review: uncertain extractions are flagged for human validation, enabling efficient review workflows.
  • Webhook & Async Processing: production-grade processing with webhook or polling to deliver results.
  • Real-world use: process invoices and receipts to auto-fill accounting records.

Quick Start

Use the DeepRead API to process a document and extract structured data. Sign up and obtain an API key from https://www.deepread.tech/dashboard. Export your API key: export DEEPREAD_API_KEY="sk_live_your_key_here" Process a document: curl -X POST https://api.deepread.tech/v1/process
-H "X-API-Key: $DEEPREAD_API_KEY"
-F "[email protected]"

Frequently Asked Questions about deepread

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I extract structured data from PDFs and images using AI?

To extract structured data from PDFs and images, use an AI-native OCR API that processes documents like invoices and contracts, returning clean JSON with field values and per-field confidence scores for efficient review.

How does confidence scoring work for OCR document processing?

Confidence scoring evaluates the certainty of extracted text and outputs a human-in-the-loop flag for uncertain fields, enabling efficient manual validation workflows for low-confidence data extractions.

Do I need an API key to process invoices and receipts into structured data?

Yes, you need to obtain an API key from the provider's dashboard and export it as an environment variable to authenticate HTTP requests when processing invoices and receipts into structured data.

Can I use webhooks for asynchronous OCR processing of large documents?

Yes, asynchronous OCR processing supports both webhooks and polling to deliver results, allowing your application to handle large document processing workflows without blocking execution.

What is the best way to automate data entry from contracts and receipts?

Automating data entry from contracts and receipts is best achieved by using a multi-model OCR API that converts documents directly into machine-readable structured JSON data, significantly reducing manual input effort.

How do I handle uncertain extractions when converting PDFs to JSON?

When converting PDFs to JSON, uncertain extractions are automatically flagged with a human-in-the-loop review flag, allowing you to isolate and manually validate only the low-confidence field values.