dataclaw

Export conversation histories from AI coding tools to Hugging Face datasets.

Updated Feb 26, 2026
One-click install
npx skills add https://github.com/MoRen9527/TriMetaverse --skill dataclaw
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: dataclaw
Source: https://github.com/MoRen9527/TriMetaverse/tree/main/reference/dataclaw/docs
Command: npx skills add https://github.com/MoRen9527/TriMetaverse --skill dataclaw

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires dataclaw, huggingface-cli, pip, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill automates the process of exporting conversation histories from various AI coding tools (Claude, Codex, Gemini CLI, OpenCode, OpenClaw) and preparing them for upload to Hugging Face, including PII auditing and redaction.

Core Features & Use Cases

  • Export Conversations: Securely export chat logs from supported AI platforms.
  • Hugging Face Integration: Streamline the process of uploading datasets to Hugging Face repositories.
  • PII Auditing & Redaction: Automatically scan exports for Personally Identifiable Information and sensitive data, with options for manual review and redaction.
  • Use Case: A developer wants to share their debugging sessions with a specific AI model on Hugging Face for community review. They use DataClaw to export their sessions, review them for any accidental PII, redact sensitive information, and then push the cleaned dataset.

Quick Start

Run dataclaw prep --source all to discover projects and check Hugging Face authentication.

Frequently Asked Questions about dataclaw

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

FAQPage Schema
How do I export conversation histories from Claude and Gemini CLI to Hugging Face?

Export conversation histories from Claude, Codex, Gemini CLI, OpenCode, and OpenClaw to Hugging Face using DataClaw. It automates dataset preparation, PII auditing, and redaction through a guided CLI workflow for secure uploads.

How does PII auditing and redaction work when uploading datasets to Hugging Face?

PII auditing scans conversation exports for sensitive data before uploading datasets to Hugging Face. DataClaw identifies personally identifiable information, allowing manual review and redaction through a staged workflow with user confirmation.

Can I use Hugging Face CLI to push exported AI chat logs as a dataset?

Yes, DataClaw integrates with Hugging Face CLI to push exported AI chat logs. It manages dataset preparation and authentication, streamlining the upload of conversation histories from supported platforms to Hugging Face repositories.

What is the best way to scan AI coding tool conversations for sensitive data before sharing?

Scanning AI coding tool conversations for sensitive data is best handled by DataClaw's PII auditing feature. It automatically detects personally identifiable information in chat logs from Claude, Codex, and Gemini CLI, enabling redaction before sharing.

Does the conversation export workflow support staged operations for reviewing redacted data?

Yes, the conversation export workflow supports staged export and push operations. DataClaw separates dataset preparation, PII redaction, and the final Hugging Face upload, requiring user confirmation before pushing sensitive data.

Why use DataClaw for dataset management instead of manually exporting chat logs?

DataClaw automates dataset management by extracting chat logs from multiple AI platforms and preparing them for Hugging Face. It eliminates manual file handling and adds automated PII auditing to ensure safe dataset sharing.