openclaw-security-audit

To run a security audit on your OpenClaw config, request the audit from OpenClaw and it will analyze the default path ~/.openclaw/openclaw.json. It performs 15+ security checks locally and outputs a markdown report with severity ratings and actionable fixes.

No, the OpenClaw security audit operates entirely locally without requiring external API calls or keys. It leverages your existing local LLM configuration to analyze the config file, ensuring sensitive data remains private and no secrets are transmitted externally.

Yes, you can audit configuration risks without exposing secrets because the audit strips sensitive values from the analysis. It reports only metadata-level findings, categorizing security risks by severity while keeping your actual credentials hidden from the LLM processing.

The OpenClaw security audit remediation report includes findings from 15+ security checks, categorized by severity levels. It provides a structured markdown roadmap with actionable fixes for each identified configuration risk, ensuring you know exactly what to remediate.

No, you do not need a specific LLM model to check OpenClaw config security risks. The audit works with whichever LLM you already use, including Opus, GPT, Gemini, or local models, requiring no additional external API keys or complex environment setup.

A limitation of using a local LLM for OpenClaw config security audits is that analysis quality depends on your local model's capabilities. Additionally, the audit strictly targets the ~/.openclaw/openclaw.json file, meaning risks in other system configurations or external dependencies are not covered.