ClawHub
Back to skill

Security audit

ClawBridge - Find Connections

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed lead-scouting skill that searches public web sources and prepares human-reviewed outreach drafts, with no evidence of hidden code, auto-sending, or credential misuse.

Install only if you are comfortable with a skill using your project profile to search public web/community sources and produce candidate briefs. Keep secrets out of profiles, configure any scheduled runner and vault/delivery tokens yourself, verify rate limits and avoid lists, and review every drafted message before sending.

SkillSpector

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

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The skill explicitly performs nightly web scouting across external sites and mentions vault uploads via a workspace token, but the description does not clearly warn users that their project profile, targeting criteria, and potentially derived candidate data may be transmitted to third-party services. This creates a meaningful privacy and compliance risk because users may supply sensitive business context or personal data without informed consent about where that data goes.

Vague Triggers

Medium
Confidence
90% confidence
Finding
The manifest description advertises a "high-signal scouting agent that runs nightly" and "matches your intent with real-world connections" without defining concrete activation boundaries, user-consent requirements, or scope limits. Broad activation language in a skill manifest can cause over-invocation or misuse in sensitive outreach, lead-generation, or profiling contexts, especially when paired with autonomous scheduling language like "runs nightly."

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.