A rigorous security auditing agent that performs a structured 100-iteration deep dive into codebases, infrastructure, and secrets.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install ralph-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 ralph-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 Ralph Security?
Ralph Security is a high-assurance security engineering agent designed to perform exhaustive audits over a 30-60 minute window. Built as one of the most robust Openclaw Skills, it follows a strict iterative engine that ensures deep analysis rather than superficial scanning. It adopts a senior security engineer persona, focusing on evidence-based findings and defense-in-depth principles.
This skill is unique because it doesn't just flag patterns; it verifies them. By reading actual code and checking for the presence of security libraries like jose, bcrypt, or Auth0, it provides high-confidence results. Whether you are managing a Node.js, Python, or Go project, this addition to your Openclaw Skills library automates the tedious parts of security compliance and vulnerability management.
Ralph Security Use Cases
- Running a comprehensive security audit before major production releases.
- Performing a weekly security check to ensure no new vulnerabilities have been introduced.
- Onboarding new projects by automatically detecting the stack and identifying architectural weaknesses.
- Auditing infrastructure-as-code and container configurations for misalignments with security best practices.
How Ralph Security Works
- Auto-Discovery: The agent identifies the tech stack, local vs remote git status, and infrastructure components like Docker or Kubernetes.
- Phased Execution: It progresses through five distinct phases: Reconnaissance, OWASP Top 10 Analysis, Authentication & Secrets, Infrastructure Security, and Code Quality.
- Verification Loop: For every iteration, the agent performs a specific check, verifies it against the code, and assigns a confidence level (e.g., VERIFIED or NEEDS_REVIEW).
- State Management: Progress is tracked with a visible
[SEC-X/100]counter, and results are auto-saved to a markdown report every 10 iterations. - Reporting: After 100 iterations, it generates a final security scorecard and a detailed report with specific fix recommendations for every failure found.
Ralph Security Setup
To initiate the audit, ensure the agent has access to your project root. You can trigger the process using natural language commands supported by Openclaw Skills.
# Standard command to start the audit
ralph security audit
# Targeting a specific phase or focus area
ralph security --phase 3 --focus secrets
The agent will automatically handle versioning of previous reports and begin the auto-detection phase.
Ralph Security Data Schema & Taxonomy
The skill organizes its findings in a structured Markdown format for easy consumption by both humans and other Openclaw Skills.
| File / Field | Description |
|---|---|
.ralph-report.md |
The primary audit file containing the current state and results. |
Severity |
Categorized as CRITICAL, HIGH, MEDIUM, LOW, or INFO. |
Confidence |
Indicates how certain the agent is based on code verification (VERIFIED to NEEDS_REVIEW). |
Phase Map |
A structured breakdown of 100 iterations across recon, OWASP, and infra segments. |
Ralph Security Advanced Features
- Checkpoint and Resume: If the audit is interrupted or reaches context limits, it can resume exactly where it left off using the saved state.
- Custom Depth: Adjust the audit intensity by modifying the iteration count parameter (up to 200 iterations).
- Library Awareness: Intelligently recognizes when security is handled by trusted third-party libraries, reducing false-positive failure reports.
- Automated Scorecard: Generates a high-level security health score as part of the final report for quick stakeholder review.
SKILL.md
Loading
Related Openclaw Skills
Ralph Quick Security Auditor
A high-speed security auditor that performs a 10-step spot-check for vulnerabilities in under 10 minutes.
Humanizer Enhanced
A professional-grade content refinement tool that detects and eliminates 34 distinct AI writing patterns while injecting personality and soul into drafts.
NadirClaw
An open-source LLM router that classifies prompts in milliseconds to slash API costs by routing to the most efficient model available.
Zero Trust Security Protocol
A security-first behavioral framework that enforces strict verification for all external AI agent operations and sensitive data handling.
Ralph Ultra
A high-intensity, 1,000-iteration security audit engine designed for deep-dive code reviews, compliance prep, and infrastructure hardening.
Researcher v2.0: USACF Research Generator
A sophisticated engine that converts rough research topics into fully-configured, multi-agent USACF swarm protocols for deep analysis.