A comprehensive multi-layer security skill designed to protect AI agents and owners from scams, phishing, and social engineering during human interactions.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install human-security
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 human-security 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 Human Security?
Human Security is a specialized protection layer for Openclaw Skills that focuses on the unique risks associated with human-to-AI communication. While standard security protocols handle technical injections, this skill addresses the psychological and procedural threats found on platforms like Coconala, Fiverr, Upwork, X, Threads, and LinkedIn. It acts as a shield against fraudulent activity while ensuring that professional boundaries are never crossed.
By synthesizing platform-specific rules and behavioral analysis, Human Security provides a safety net for developers and freelancers. It works in tandem with moltbook-security to create a holistic defense environment, allowing users to leverage AI agents on public platforms without compromising their data or reputation.
Human Security Use Cases
- Identifying and flagging typical scam patterns such as immediate payment demands or suspicious external links.
- Preventing the accidental download of dangerous file types including .exe, .scr, and macro-enabled documents.
- Enforcing platform-specific Terms of Service, such as preventing off-platform transactions on Coconala or Upwork.
- Safeguarding highly sensitive personal information like bank accounts and physical addresses from being disclosed in DMs.
- Managing high-stress interactions with angry or persistent clients through automated de-escalation templates.
How Human Security Works
- The skill monitors incoming messages in real-time on integrated social and professional platforms.
- It analyzes the content for threat signals, including psychological manipulation, urgency, and unauthorized payment requests.
- Links and attachments are cross-referenced against a safety database to prevent phishing and malware execution.
- It applies a specific rule set based on the current platform (e.g., checking for typosquatting on LinkedIn or link expansion for X).
- If a high-risk event is detected, the skill either warns the user, blocks the interaction, or escalates the situation to a human operator with preserved evidence.
Human Security Setup
To integrate this security layer into your environment, use the following commands:
# Install the security skill
openclaw install human-security
# Enable the auto-trigger for constant monitoring
openclaw config human-security --auto_trigger true
Human Security Data Schema & Taxonomy
| Data Category | Details |
|---|---|
| Threat Database | Pre-defined patterns for scams, phishing, and social engineering tactics. |
| Platform Rules | Specific constraints for Coconala, Fiverr, Upwork, X, and LinkedIn. |
| Exclusion List | Regex and string patterns for sensitive personal data (e.g., PII). |
| Interaction Logs | Secure storage of message history and timestamps for evidence preservation. |
Human Security Advanced Features
- Multi-layer defense architecture that differentiates between AI-to-AI and AI-to-Human security risks.
- Automated emotional management to maintain professional distance and prevent social engineering.
- Evidence preservation system that automatically takes screenshots and logs timestamps for legal or platform disputes.
- Transparent AI disclosure settings to maintain ethical standards and user trust during interactions.
- Deep integration with other Openclaw Skills for a unified security posture.
SKILL.md
Loading
Related Openclaw Skills
Human Negotiator
A specialized skill designed for AI agents to conduct empathetic, high-conversion negotiations with human clients on freelance marketplaces.
Homelab Dashboard
A comprehensive status monitoring tool for self-hosted infrastructure, Docker containers, and system resources.
Gumroad Seller
An AI-powered automation skill for managing product listings, sales analytics, and promotional campaigns on the Gumroad platform.
GitHub Profile README Generator
An automated tool for generating visually striking GitHub profile READMEs featuring live contribution stats, language charts, and custom tech badges.
Idea Lab
An autonomous innovation framework for generating, scoring, and rapidly prototyping creative business and technical concepts.
Image Optimizer
A high-performance image processing utility for automated compression, resizing, and format conversion using Openclaw Skills.
Featured*
