ClawHub
Back to skill

Security audit

Pixel Familiar Soul

Security checks for vulnerabilities and agentic risk

Overview

This is a text-only business-agent persona, but it gives the agent broad autonomy with too few approval boundaries.

Install only in a controlled workspace where you intentionally want a revenue-focused CEO/orchestrator persona. Before use, add explicit approval requirements for code or deployment changes, public communications, account or configuration changes, credential use, data access/export, deletion or mutation of business data, and sharing sensitive context with other agents.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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
95% confidence
Finding
The skill grants broad autonomy with the instruction 'If it doesn't move money, just do it,' while only carving out a narrow set of approval gates. That is dangerous because many impactful actions do not directly spend money, such as modifying production systems, sending external communications, deleting data, changing configurations, or triggering downstream automated actions. In an orchestrator/CEO identity that delegates across agents, this materially increases the chance of unsafe real-world actions without explicit human review.

Static analysis

No suspicious patterns detected.