ClawKit Logo
ClawKitReliability Toolkit
Back to Registry
Official Verified productivity Safety 4/5

PersonalDataHub

Pull personal data (emails, issues) and propose outbound actions (drafts, replies) through the PersonalDataHub access control gateway. Data is filtered, redacted, and shaped by the owner's policy before reaching the agent.

Why use this skill?

Use PersonalDataHub to connect OpenClaw to Gmail and GitHub. Benefit from privacy-first data redaction, secure MCP integration, and automated email drafting.

skill-install — Terminal

Install via CLI (Recommended)

clawhub install openclaw/skills/skills/haojian/personaldatahub
Or

What This Skill Does

The PersonalDataHub skill acts as a secure, policy-driven bridge between the OpenClaw agent and your personal data silos, including Gmail and GitHub. Unlike raw data integration, this skill routes all requests through an access control gateway that ensures data is filtered, redacted, and shaped according to your specific privacy preferences before the agent ever sees it. By using the MCP (Model Context Protocol), it dynamically discovers available tools only after you have successfully authenticated with the required OAuth services, ensuring a zero-trust approach to your personal information.

Installation

To integrate PersonalDataHub, use the OpenClaw installation utility. Run the following command in your terminal:

clawhub install openclaw/skills/skills/haojian/personaldatahub

After installation, configure the MCP server by updating your Claude Code settings file located at .claude/settings.json. Add the following configuration to enable native tool discovery:

{
  "mcpServers": {
    "personaldatahub": {
      "command": "npx",
      "args": ["pdh", "mcp"]
    }
  }
}

Restart your agent to complete the registration process.

Use Cases

This skill is perfect for professionals who want to automate administrative overhead without sacrificing data security. Common use cases include: summarizing complex technical issue threads on GitHub to prepare for daily standups, drafting follow-up emails for unread messages, or managing communication workflows where the agent provides a first-pass draft that you then approve and refine. It essentially turns your messy inbox and issue trackers into structured, actionable insights.

Example Prompts

  1. "Check my inbox for any urgent emails from my manager regarding the Q4 project, and draft a short reply acknowledging receipt."
  2. "Search GitHub for issues assigned to me in the core-engine repository that have been open for more than 5 days."
  3. "Summarize the recent email thread about the website redesign and draft a response suggesting a meeting for next Tuesday."

Tips & Limitations

  • Safety First: All outbound communication tools (drafting, sending, replying) stage the content for your manual review. The agent cannot send an email without your explicit confirmation.
  • Privacy Policy: Take time to configure your gateway policy. If you do not want the agent to see specific contact names or PII, ensure those fields are redacted in the gateway settings.
  • Connectivity: Tools are only visible when OAuth tokens are active. If a tool is missing from your list, re-authorize the relevant service in your connection manager.

Metadata

Author@haojian
Stars2387
Views0
Updated2026-03-09
View Author Profile
AI Skill Finder

Not sure this is the right skill?

Describe what you want to build — we'll match you to the best skill from 16,000+ options.

Find the right skill
Add to Configuration

Paste this into your clawhub.json to enable this plugin.

{
  "plugins": {
    "official-haojian-personaldatahub": {
      "enabled": true,
      "auto_update": true
    }
  }
}

Tags(AI)

#gmail#github#privacy#automation#mcp
Safety Score: 4/5

Flags: data-collection, external-api