🔒 Transparent by design

Security & Privacy
No surprises. Ever.

AI agents that touch your files and accounts deserve scrutiny. Here's exactly what our skills do, what stays on your machine, and how to lock everything down.

🏠 What Stays on Your Machine

100% Local

Your Memory Files

MEMORY.md, daily memory logs, all agent context files. These never leave your machine. They're plain text files you can read, edit, or delete anytime.

100% Local

Cron Jobs & Schedules

All cron job definitions, schedules, and execution logs are stored locally in your OpenClaw config. No cloud scheduler involved.

100% Local

Skill Files

Every skill you install lives in your ~/.openclaw/workspace/skills/ directory. You can inspect every line of code before it runs.

100% Local

API Keys & Config

Your API keys, tokens, and credentials are stored in your local openclaw.json. They are never sent to GetAgentIQ or any third party.

📡 What Leaves Your Machine

LLM API Calls Only

The only data that leaves is what you send to your AI model provider

  • Your prompts and conversation context are sent to your chosen LLM (Anthropic, OpenAI, xAI etc.)
  • Any personal information you include in prompts is subject to that provider's privacy policy
  • Web scraping skills make outbound HTTP requests to the URLs you specify
  • Email skills connect to your Gmail/inbox using OAuth — only with your explicit authorisation

GetAgentIQ never receives your prompts, responses, or personal data. We have no telemetry, no analytics, no callbacks in our skills.

🔐 How to Lock Down Access

  • Use OpenClaw's tool allowlists to restrict which tools each agent can use
  • Enable elevated permission prompts — agent must ask before any destructive action
  • Set ask=always mode for high-risk tools (exec, file deletion, external sends)
  • Review installed skills with our Skill Vetter (included in Security Pack) before trusting them
  • Run agents in sandbox mode until you trust what they're doing
  • Use audit logs to see exactly what every cron job did during its run

🛡️ Our Security Pack — £69

6 Skills

What you get with the Security Pack

  • Skill Vetter — scans any skill for malicious patterns before you install it
  • Zero-Trust Config — pre-built allowlist templates for locked-down agents
  • Audit Logger — structured logs of every tool call, cron run, and external request
  • Permission Guardian — human-in-the-loop for high-risk actions
  • Secret Scanner — detects hardcoded credentials in installed skills
  • Session Integrity Monitor — alerts on session bloat and unexpected behaviour

❓ Common Questions

Can OpenClaw read my emails without permission?
No. Email access requires you to explicitly install the Gmail skill and authorise it via Google OAuth. Without that, OpenClaw has no access to your inbox. You can revoke access anytime from your Google account security settings.
Does it send my data to third parties?
GetAgentIQ skills contain no telemetry, tracking, or data callbacks. The only external calls are to your chosen LLM provider (for AI responses) and to APIs you've explicitly configured (e.g. Gmail, Stripe, web scraping targets). We have no visibility into what your agent does.
What happens if a skill is malicious?
This is a real risk in any skill ecosystem. That's why we built the Skill Vetter — it scans for eval() abuse, hardcoded secrets, suspicious network calls, and reverse shell patterns before you run anything. Our Security Pack includes it. For any skill from unknown sources: always inspect the SKILL.md and scripts before installing.
How do I know what it's doing?
OpenClaw shows every tool call in your chat interface. The Audit Logger (Security Pack) additionally writes structured logs of every action to a local file. You can see exactly what ran, when, what it accessed, and what the output was. Nothing is hidden.
Are GetAgentIQ skill packs safe to install?
Every skill in our packs is written by us, reviewed by our Viper QA agent, and passes our own Skill Vetter scan before release. We publish release notes for every skill so you can see exactly what changed between versions. We also include a LICENSE.txt in every ZIP download.

Take Control of Your Agent's Security

The Security Pack gives you the tools to audit, restrict, and monitor every action your agent takes — with zero guesswork.

Get the Security Pack — £69
← Back to all packs