ClawHub
Back to skill

Security audit

SaaS

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed local note-taking and SaaS-operations guide with sensitive business memory, but its storage and credential boundaries are clear and purpose-aligned.

Install only if you are comfortable with the agent maintaining local SaaS business memory under ~/Clawic/data/, including account names, ARR, commitments, churn reasons, contacts, and project records. Keep that directory access-controlled, and do not paste secrets or full customer/user exports into sessions you expect the skill to remember.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The skill explicitly instructs users to record diligence outputs, restated numbers, commitments, concentration analysis, and process artifacts into shared files and project records, but it provides no data-minimization, access-control, or redaction guidance. In the context of diligence, those materials can contain sensitive customer, contract, legal, receivables, and security/compliance information, so normalizing their storage in shared locations increases the risk of unauthorized disclosure or oversharing across teams and future processes.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.