A comprehensive bilingual security auditor and real-time shield providing 68 automated checks and 11 self-healing fixes for agent environments.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install lobsterguard
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 lobsterguard 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 lobsterguard?
lobsterguard is a high-performance security framework designed to audit and protect your local environment when using Openclaw Skills. It functions as both a forensic auditor and a proactive shield, covering six critical security categories including system hardening, network safety, and the OWASP Agentic AI Top 10 vulnerabilities.
By leveraging local scripts and a dedicated gateway plugin, lobsterguard ensures that your AI agent operations remain private and secure. It intercepts potential threats like prompt injections and path traversals in real-time, providing a robust layer of defense that is essential for developers running autonomous agents on their local machines.
lobsterguard Use Cases
- Hardening local system configurations to prevent unauthorized access through Openclaw Skills.
- Real-time detection and blocking of prompt injection and command injection attacks.
- Automated remediation of common security vulnerabilities such as exposed ports or weak kernel parameters.
- Monitoring agent activity for forensic evidence of suspicious processes or file modifications.
- Managing secure migrations of agent environments from root to dedicated system users.
How lobsterguard Works
- The auditor executes 68 local checks across the system and agent configuration to identify security gaps.
- A compact report is generated, displaying failed checks, a security score, and available auto-fixes.
- For fixable issues, the skill generates a detailed remediation plan for user review.
- Upon user confirmation, lobsterguard executes hardening steps using sudo access for system-level changes.
- The verification engine runs a follow-up scan to confirm the vulnerability has been successfully mitigated.
- The Gateway Shield plugin continuously monitors incoming prompts and file changes to intercept threats before execution.
lobsterguard Setup
To install lobsterguard and integrate it with your Openclaw Skills, run the following commands in your terminal:
git clone https://github.com/jarb02/lobsterguard.git
cd lobsterguard
chmod +x install.sh
./install.sh
Ensure you configure your TELEGRAM_BOT_TOKEN and TELEGRAM_CHAT_ID in your environment variables to receive real-time security alerts.
lobsterguard Data Schema & Taxonomy
lobsterguard organizes its security data and scripts within the standard agent directory structure:
| Directory/File | Purpose |
|---|---|
scripts/ |
Contains Python logic for scans, fixes, and Telegram utilities |
extension/ |
Houses the Gateway Shield plugin for real-time interception |
systemd/ |
Configuration for automated scanning and quarantine services |
reports/ |
Local storage for scan history and security audits |
quarantine/ |
Isolated directory for suspicious skills or malicious files |
lobsterguard Advanced Features
- Gateway Shield plugin that intercepts 31 distinct threat patterns in real-time.
- Automated systemd integration for background scanning and persistent quarantine monitoring.
- Dedicated OWASP Agentic AI Top 10 coverage to protect against RAG poisoning and rogue agents.
- Telegram integration supporting 16 remote commands for scanning and firewall management.
- Non-intrusive forensic detection that analyzes logs and processes without external API calls.
SKILL.md
Loading
Related Openclaw Skills
ADHD Daily Planner
A bullet journal-style planning system designed to help ADHD brains manage tasks using energy-based swim lanes instead of rigid time blocks.
ADHD Body Doubling
A high-accountability productivity skill designed to provide neurodivergent founders with punk-style body doubling, micro-step breakdown protocols, and frequent check-ins.
Clawdbot Documentation Expert
An AI-powered guide designed to help developers navigate, configure, and troubleshoot Clawdbot environments efficiently.
Apple Developer Documentation Skill
A bridge for AI agents to search and retrieve official Apple Developer Documentation and API references.
4claw
A specialized, moderated imageboard protocol that allows AI agents to create threads, post replies, and participate in community-driven boards.
Mermaid Diagrams Skill
The Mermaid Diagrams Skill enables the programmatic generation of visual charts and diagrams directly from text-based Mermaid syntax.
METADATA
Featured*
