SecureClaw has 51 audit checks + 15 behavioral rules for OpenClaw agents. ATR (Agent Threat Rules) maintains 108 regex detection rules covering a different layer — MCP tool poisoning, SKILL.md supply chain attacks, credential exfiltration, and prompt injection patterns.
They complement each other: SecureClaw does runtime hardening + behavioral rules, ATR does pattern-based threat detection.
Cisco AI Defense ships 34 ATR rules in production. 53K+ skills scanned, 0% FP on clean content.
Would you be open to a PR that integrates ATR detection patterns alongside SecureClaw's existing rules? Happy to match your format.
SecureClaw has 51 audit checks + 15 behavioral rules for OpenClaw agents. ATR (Agent Threat Rules) maintains 108 regex detection rules covering a different layer — MCP tool poisoning, SKILL.md supply chain attacks, credential exfiltration, and prompt injection patterns.
They complement each other: SecureClaw does runtime hardening + behavioral rules, ATR does pattern-based threat detection.
Cisco AI Defense ships 34 ATR rules in production. 53K+ skills scanned, 0% FP on clean content.
Would you be open to a PR that integrates ATR detection patterns alongside SecureClaw's existing rules? Happy to match your format.