A specialized AI skill that extracts structured JSON data from pay stubs and earnings statements with built-in confidence scoring for precise income verification.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install deepread-pay-stubs
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install deepread-pay-stubs using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is DeepRead Pay Stubs?
The DeepRead Pay Stubs skill is one of the most powerful Openclaw Skills for automating financial data processing and document intelligence. It enables AI agents to effortlessly parse pay stubs, wage statements, and earnings summaries into clean, typed JSON format. Instead of dealing with unformatted text, you get organized data fields including employee information, exact pay periods, net and gross pay, itemized deductions, and YTD aggregates.
Designed for risk-sensitive workflows, this tool provides a granular "needs_review" boolean key for every single extracted field. This empowers developers to flag low-confidence outputs automatically, ensuring high accuracy for financial decision-making. Utilizing this skill within your suite of Openclaw Skills simplifies building robust, automated document-processing applications.
DeepRead Pay Stubs Use Cases
- Lending and Mortgage Automation: Streamline applicant income verification directly from submitted payroll documents.
- Tenant and Rental Screening: Instantly compute income-to-rent ratios to qualify applicants.
- HR & Payroll Reconciliations: Match physical or digital employee pay stubs against central payroll system ledgers.
- Gig Economy & Benefits Eligibility: Process non-standard income records and calculate annualized earnings dynamically.
How DeepRead Pay Stubs Works
- The AI agent detects an incoming pay stub document (PDF or image format) and uploads it securely.
- The skill routes the document to the DeepRead API endpoint (
https://api.deepread.tech). - The agent polls the API for processing completion status.
- DeepRead extracts the payload fields, evaluates extraction confidence, and formats the output into structured JSON.
- The finalized payload, complete with field-level confidence indicators and localized page information, is returned to the agent.
DeepRead Pay Stubs Setup
To begin using this skill, you must obtain an API key and set it in your environment variables:
# 1. Open the DeepRead dashboard and register for a free account (includes 2,000 pages/month)
open "https://www.deepread.tech/dashboard/?utm_source=clawhub"
# 2. Export your secret API key
export DEEPREAD_API_KEY="sk_live_your_key_here"
DeepRead Pay Stubs Data Schema & Taxonomy
JSON Extraction Schema
Each successful extraction maps pay stub fields into a standardized schema structure:
| Field Name | Type | Description |
|---|---|---|
employer_name |
String | Name of the issuing employer |
employee_name |
String | Name of the payee/employee |
pay_period_start |
String | Start date of the pay cycle (YYYY-MM-DD) |
pay_period_end |
String | End date of the pay cycle (YYYY-MM-DD) |
pay_date |
String/Null | Date the payment was issued (YYYY-MM-DD) |
pay_frequency |
String/Null | Frequency of payment (e.g., Weekly, Biweekly, Semimonthly, Monthly) |
gross_pay |
Number | Gross earnings before taxes and deductions |
net_pay |
Number | Net take-home earnings |
ytd_gross |
Number/Null | Year-to-date cumulative gross earnings |
ytd_net |
Number/Null | Year-to-date cumulative net earnings |
deductions |
Array | List of itemized deductions with type and amount |
Sample Response Payload
{
"schema_version": "dp02",
"status": "completed",
"extraction": {
"fields": [
{
"key": "employer_name",
"value": "Acme Corp",
"needs_review": false,
"location": { "page": 1 }
},
{
"key": "gross_pay",
"value": 3520.0,
"needs_review": false,
"location": { "page": 1 }
}
]
}
}
DeepRead Pay Stubs Advanced Features
- Per-Field Confidence (needs_review): Minimizes errors by flagging values that require human oversight.
- PII Redaction Support: Pairs natively with other Openclaw Skills such as
deepread-piito ensure sensitive client details are redacted prior to external transmission. - Custom Parsing Blueprints: Configure layout templates in the DeepRead Optimizer dashboard to guarantee layout consistency.
- Programmatic Verification Logic: Easily compute annualized incomes or calculate mathematical balances (e.g.,
net_pay + deductions ≈ gross_pay) using code integration.
SKILL.md
Loading
Related Openclaw Skills
World Cup 2026 Quantitative Football Analysis
An advanced, search-calibrated quantitative football prediction and tournament simulation engine designed specifically for the 2026 World Cup.
The Boys in the Boat Team Dynamics & Perseverance Toolkit
An interactive AI agent skill that translates the historic team synchronization, perseverance, and leadership lessons from Daniel James Brown's The Boys in the Boat into actionable modern frameworks.
Google Maps Scraper
A powerful lead generation skill that extracts business names, websites, emails, phone numbers, and reviews directly from Google Maps into organized CSV files.
The Big Leap: Conquer Self-Sabotage & Find Your Zone of Genius
An AI-powered cognitive behavioral toolkit and peak performance coach based on Gay Hendricks' framework to help you identify self-sabotage and operate in your Zone of Genius.
MagicHaqi Pet Master
A specialized browser automation skill that enables AI agents to co-parent a MagicHaqi virtual pet by interacting directly with the game interface, executing daily care workflows, and managing pet stats.
Arabic Localization Reviewer Skill
An advanced quality assurance and localization skill designed to review Arabic and bilingual Arabic/English content, RTL layout integrity, and translation accuracy.
Featured*
