ClawHub
Back to skill

Security audit

HaluCatch / 捕幻

Security checks across malware telemetry and agentic risk

Overview

HaluCatch is a local skill-audit tool that reads a user-selected skill folder and writes Markdown reports, with no evidence of hidden network access, credential use, destructive behavior, or persistence beyond report files.

Install only if you are comfortable letting it recursively read the specific skill folder you name and create local Markdown audit reports. Do not use a home directory or unrelated private project as the target, and if you use --output-dir, make sure the destination is intentional.

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)

Vague Triggers

Medium
Confidence
93% confidence
Finding
The trigger examples include broad natural-language phrases such as asking to 'audit a skill' or 'quickly scan for obvious issues,' which are common requests that could accidentally invoke this skill in contexts where the user did not intend file scanning, script execution, or report generation. Because the skill also has Read/Write/Bash capabilities, unintended activation could lead to unnecessary access to local directories and file creation under `reports/`.

Natural-Language Policy Violations

Medium
Confidence
88% confidence
Finding
The skill instructs the AI to infer the user's language and automatically force Chinese output for Chinese users without explicit consent. This is not a classic security exploit, but it can override user preference, cause confusing behavior, and make downstream review or logging inconsistent if a user expects another language.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.