ClawHub

Defluff

Security checks across malware telemetry and agentic risk

Overview

This appears to be a purpose-aligned email analysis skill, with minor scoping and language-preference caveats rather than evidence of hidden or unsafe behavior.

Install this if you want an agent to analyze email text for scam, authorship, or prompt-injection signals. Be careful with private emails, and ask explicitly for the analysis type and output language you want so the skill does not over-apply to ordinary summaries.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Vague Triggers

Medium
Confidence
93% confidence
Finding
The trigger criteria are broad enough to match ordinary email summarization and triage requests, so the skill may activate when the user did not intend to perform AI-authorship detection or scam classification. That can cause misrouting to a specialized skill, privacy over-processing of email content, and misleading outputs in contexts where a plain summary was requested.

Natural-Language Policy Violations

Medium
Confidence
88% confidence
Finding
Forcing the output to mirror the email's language without user opt-in can override user preferences, reduce comprehension, and create policy or accessibility issues when the user expects analysis in a different language. In a security workflow, this may also hinder review if analysts cannot read the mirrored language, increasing the chance that phishing or fraud indicators are missed.

VirusTotal

61/61 vendors flagged this skill as clean.

View on VirusTotal