Openclaw Shield 1.0.3

Enterprise AI security scanner combining static code analysis, runtime guards, ClamAV integration, and tamper-proof audit logging to detect threats and enfor...

This skill appears to be a legitimate security scanner, but it asks you (or the agent) to fetch and run code from an external GitHub repository and references integrations (ClamAV, Telegram, cron) without declaring required binaries or credentials. Before installing or running it: 1) review the referenced GitHub repo source (src/scanner.py and deployment scripts) and a pinned commit or tag; 2) confirm what environment variables and host binaries (ClamAV, Telegram bot token) it will actually need; 3) verify there are checksums or signed releases for any code you fetch; 4) run the scanner in an isolated sandbox or non-production account first; 5) audit any cron install scripts and allowlist enforcement code to ensure it won't alter unrelated system settings; and 6) note the metadata mismatch (owner IDs/published version/status claims) and ask the publisher for clarification if you plan to trust it in a production environment.

Type: OpenClaw Skill Name: openclaw-shield-1-0-3 Version: 1.0.0 The skill bundle lacks the actual source code (src/scanner.py) it claims to provide, instead instructing the AI agent to download and execute code from an external GitHub repository (github.com/pfaria32/OpenClaw-Shield-Security), which is a significant supply chain risk. The inclusion of a highly defensive SECURITY.md file, which preemptively justifies why the tool might be flagged as malicious, is a common social engineering tactic used to bypass security scrutiny. Furthermore, SKILL.md contains instructions that claim the tool is already 'Deployed' on a specific instance (clawdbot-toronto), which may be an attempt to manipulate the agent's trust or environment state.