A local-first, read-only security auditing tool designed to assess and harden OpenClaw instance configurations against leaks and misconfigurations.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install m78armor-security-check
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 m78armor-security-check 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 m78armor Security Check?
m78armor Security Check is a specialized utility within the Openclaw Skills ecosystem focused on maintaining the integrity and privacy of your local AI agent environment. It performs a comprehensive, read-only review of your configuration files to identify permission gaps, risky default settings, and configuration drift that could lead to unauthorized access or data exposure. Designed with a privacy-first approach, it operates entirely offline and never uploads local data, ensuring your environment remains secure and confidential.
By comparing your current setup against a hardened security baseline, this skill provides actionable, evidence-based insights. It translates complex security risks into plain language, explaining how specific settings could be abused and providing clear, practical steps for manual remediation. It is an essential post-installation or post-upgrade tool for developers who prioritize the safety of their AI agent workflows.
m78armor Security Check Use Cases
- Validating security baselines immediately after a new installation or version upgrade.
- Identifying configuration drift where local settings have deviated from secure defaults over time.
- Auditing file permissions and exposure gaps to prevent local data leaks.
- Assessing the risk profile of an OpenClaw instance before deploying sensitive AI tasks.
- Generating machine-readable security reports for local environment documentation.
How m78armor Security Check Works
- The skill detects the operator's preferred language (English or Simplified Chinese) to provide localized feedback.
- It executes a bundled Node.js script that targets the local OpenClaw configuration directory.
- The engine performs a read-only scan of configuration files, looking for known misconfigurations and high-risk defaults.
- It analyzes identified findings to determine potential abuse paths and operational impact.
- The skill outputs a factual, evidence-based report detailing findings, risk explanations, and concise remediation guidance.
m78armor Security Check Setup
To use this security check within the Openclaw Skills framework, ensure you have Node.js installed on your system.
# Check if Node.js is installed
node -v
# Run the security assessment in English
node scripts/m78armor-lite.js --lang en
# Run the security assessment in Chinese
node scripts/m78armor-lite.js --lang zh
m78armor Security Check Data Schema & Taxonomy
The skill organizes its security audit data based on logical risk domains. When the --json flag is used, it produces a structured schema for integration with other tools:
| Field | Description |
|---|---|
findings |
A list of identified security concerns or misconfigurations. |
severity |
The categorized risk level (e.g., info, warning, high). |
description |
A plain-language explanation of the configuration status. |
remediation |
Specific manual steps required to harden the setting. |
path |
The local file path or configuration key where the finding was detected. |
m78armor Security Check Advanced Features
- Local-first architecture ensuring zero data exfiltration during security audits.
- Bilingual support with specialized logic for English and Simplified Chinese environments.
- Custom configuration path targeting using the
--configargument for non-standard installs. - Machine-readable JSON output mode for automated reporting and CI/CD integration.
- Read-only guardrails to prevent accidental system changes during assessment.
SKILL.md
Loading
Related Openclaw Skills
ALTER Identity Server
A neutral identity infrastructure for the AI economy that verifies humans behind agents and enables data-driven earnings.
Wenshu Search General Skill
An automated browser-based skill for searching, extracting, and cleaning full-text legal judgment documents from China Judgments Online.
Video Frame Extraction
A robust technical skill for AI agents to extract high-quality image frames from various video formats using OpenCV for automated media processing.
gTTS (Google Text-to-Speech)
A powerful Python-based skill that leverages Google's Text-to-Speech API to transform written content into spoken audio without requiring an API key.
TuhuCar Knowledge Assistant
A professional automotive knowledge gateway that provides precise car maintenance, service intervals, and technical specifications via the TuhuCar CLI.
Stanley Druckenmiller Cognitive OS
A sophisticated macro investing cognitive operating system based on the legendary 30-year trading framework of Stanley Druckenmiller.
Featured*
