A systematic workflow for verifying numerical data against original research papers to maintain 100% traceable provenance.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install data-source-verification
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 data-source-verification 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 Data Source Verification?
The Data Source Verification skill is an essential framework for researchers and developers using Openclaw Skills to ensure the absolute integrity of their datasets. It establishes a rigorous Data Provenance Chain that links every numerical value—whether in a CSV, JSON, or LaTeX manuscript—directly back to a specific location within a source PDF (such as a table, figure, or text passage).
By implementing this skill, users can eliminate data hallucinations and misattributions common in automated data collection. It provides a structured method for managing citation sources, extracting metadata, and performing deep audits to confirm that every data point is verified, approximate, or appropriately flagged for manual review.
Data Source Verification Use Cases
- Building high-fidelity datasets from scientific literature (CSV, JSON, SQL).
- Populating academic tables or plots with values derived from multiple peer-reviewed papers.
- Performing data integrity audits on existing datasets before publication or model training.
- Cross-referencing experimental results with DFT-calculated or simulated data points.
How Data Source Verification Works
- Initialize a structured directory for citation sources and metadata.
- Add original source PDFs and supplementary information to the local environment.
- Create structured CITATION.md files to capture DOIs, titles, and extracted values.
- Extract data with explicit provenance, recording the exact table, figure, or page number.
- Classify data points based on verification status (e.g., VERIFIED, UNVERIFIED, or MISATTRIBUTED).
- Execute an audit workflow to scan for missing PDFs or unconfirmed values.
- Export a final Data Provenance Summary table for transparency and peer review.
Data Source Verification Setup
To begin using this framework within your Openclaw Skills environment, initialize your project structure:
mkdir -p Citation_Sources/AuthorYear_Journal_Title
touch Citation_Sources/AuthorYear_Journal_Title/CITATION.md
Ensure you have a PDF viewer or OCR capabilities enabled if you are using an AI agent to assist in the extraction process.
Data Source Verification Data Schema & Taxonomy
The skill organizes information using a standardized provenance model to ensure every value is auditable:
| Field | Description | Example |
|---|---|---|
| Value | The actual numerical data | 0.69 |
| Source_Paper | Primary author and year | Cheng 2021 |
| Source_Location | Exact position in PDF | Table 2, row 3 |
| Data_Type | Origin of the data | Experimental, DFT, or Derived |
| Verified | Boolean or status flag | YES, NEEDS CONFIRM, MISATTRIBUTED |
Data can be exported in both CSV and JSON formats, ensuring compatibility with most data science pipelines.
Data Source Verification Advanced Features
- Multi-agent provenance tracking for collaborative dataset building.
- Automated discrepancy detection between conflicting research sources.
- Support for structured metadata taxonomy using Citation_Sources directory nesting.
- Detailed audit reporting that generates a health score for your entire dataset.
- Custom status labels for paywalled content and manual verification triggers.
SKILL.md
Loading
Related Openclaw Skills
MBTI Fortune Teller
An innovative divination tool that uses MBTI cognitive function stacks to provide structured, personality-driven answers to any question.
Capacities Lookup
A specialized skill for finding and deep-linking Capacities objects via real-time API lookup and metadata enrichment.
Article Creation Workflow
A full-loop AI workflow that transforms product keywords into a strategic library of high-converting English marketing articles.
AI News Pipeline
A high-frequency automated workflow for capturing international and domestic AI news, generating summaries, and producing detailed reports in Excel and Word.
Abfallkalender RV
A technical tool for AI agents to retrieve and process waste disposal schedules for the Ravensburg district via official web sources.
Cleanup Reporter
A local file system scanner that generates detailed reports on large directories and duplicate files to help you reclaim disk space.
Featured*
