ClawHub
Back to skill

Security audit

Customer Support Command Center

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent customer-support playbook, but it gives agents under-scoped authority around customer data, account locking, refunds, and automated customer replies.

Install only if you will keep customer-data export, deletion, account locking, refunds, credits, and outbound replies behind your existing support system permissions and human approval rules. Treat this as an operations template, not as permission for an agent to autonomously change accounts or handle privacy requests.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • YARA SignaturesMalware Match, Webshell Match, Cryptominer Match
  • 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 (3)

YARA rule 'agent_skill_mcp_tool_poisoning_metadata': MCP/tool metadata poisoning indicators in tool schemas or skill manifests [agent_skills]

High
Category
YARA Match
Confidence
80% confidence
Finding

YARA rule matched a hack tool or exploit indicator (offensive tools, reconnaissance, privilege escalation, or exploit frameworks).

Content

Scanner excerpt · SKILL.md (reported line 3)May include surrounding context.

md
---
name: Customer Support Command Center
description: "Enterprise-grade customer support system: ticket triage, response templates, escalation workflows, CSAT tracking, knowledge base management, and churn prevention. Turns your AI agent into a support team lead."
metadata: {"clawdbot":{"emoji":"🎯","os":["linux","darwin","win32"]}}
---

# Customer Support Command Center

You are a customer support operations agent. You handle ticket triage, craft responses, manage escalations, track satisfaction, maintain the knowledge base, and prevent churn. You don't just answer questions — you run the entire support function.

---

## 1. Tic

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The README advertises automation features such as auto-routing, canned responses, and confidence-gated auto-replies without any visible warning about reviewing outputs, protecting customer data, or limiting outbound actions. In a customer support context, this can lead users to deploy autonomous messaging or handling of sensitive customer information without adequate safeguards, increasing the risk of privacy violations, incorrect communications, and operational harm.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
85% confidence
Finding

The skill instructs the agent to fulfill GDPR/privacy requests and lock accounts, but it does not require strong safeguards such as verified authorization, human approval, audit logging, scope checks, or explicit confirmation before taking actions that affect personal data or customer access. In a support-agent context, this creates a real risk of unauthorized data disclosure, deletion, or account denial-of-service if the agent follows the playbook too directly.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.