ClawHub
Back to skill

Security audit

Openclaw Skill

Security checks across malware telemetry and agentic risk

Overview

Phishguard is a disclosed phishing-link checker that runs local URL analysis and fetches public blocklist shards from GitHub, with no evidence of credential theft, hidden exfiltration, destructive behavior, or privilege escalation.

Install if you are comfortable with the skill running python3, fetching public blocklist shard files from GitHub, and storing a small local cache. Treat “safe” results as advisory rather than a guarantee, especially because the checker uses heuristics and allowlists as well as blocklist data.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

Description-Behavior Mismatch

Medium
Confidence
94% confidence
Finding
The skill metadata advertises real-time checks against 38 threat intelligence sources, but this file only retrieves a single GitHub-hosted shard set and applies local heuristics. In a phishing-detection skill, this discrepancy is security-relevant because users may overtrust negative results and click malicious links under a false sense of protection.

Vague Triggers

Medium
Confidence
82% confidence
Finding
The instruction to proactively scan URLs whenever the context 'suggests they want safety advice' is ambiguous and can cause the agent to process links from broader user messages without clear, explicit consent. In a security tool, this increases the chance of over-collection, unintended network activity, and surprising behavior, especially when scanning triggers local execution and remote shard fetches.

VirusTotal

54/54 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.