企服客户管理

• Outcome: Review
• Latest audit: Jun 8, 2026 at 1:30 AM UTC
• Version:
  1.0.0

Overview

This customer-management skill is coherent, but it handles real customer records and includes an embedded WeCom webhook plus scheduled outbound risk-report pushing without enough consent or data-control boundaries.

Review before installing. Use only in an authorized business environment, remove and rotate the exposed WeCom webhook, keep local-only mode as the default, and require explicit approval plus redaction before sending tenant, contact, payment, or risk data to WeCom or Tencent Docs.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (4)

Context-Inappropriate Capability

Medium

• Confidence: 88% confidence
• Finding: The skill documentation includes optional pushing to WeCom groups and generating Tencent Docs, which introduces outbound data-sharing capabilities on top of local Excel processing. Because the workbook contains real customer data and the feature is only loosely described, operators may unknowingly transmit sensitive tenant and risk information to external systems without proper authorization or minimization.

Context-Inappropriate Capability

High

• Confidence: 97% confidence
• Finding: The configuration hardcodes a WeCom webhook URL, enabling direct outbound delivery of customer-related reports to an external messaging endpoint. Embedding a live webhook in skill documentation/config creates a real exfiltration path for sensitive business and tenant data and also risks unauthorized use of the webhook itself if the file is shared or committed.

Missing User Warnings

High

• Confidence: 93% confidence
• Finding: The skill describes automated generation and push of customer risk reports to enterprise messaging channels, but it does not provide an explicit user-facing warning that sensitive customer data may leave the local workbook environment. In this context, the data includes real tenant identities, contact details, payment status, and risk classifications, making silent external dissemination particularly dangerous.

Missing User Warnings

High

• Confidence: 98% confidence
• Finding: The configuration both exposes a specific external WeCom webhook and lacks any clear warning that customer data may be transmitted over the network. Given the skill processes real business records from Excel, this creates a direct and non-transparent channel for sensitive data disclosure outside the local system boundary.

VirusTotal

64/64 vendors flagged this skill as clean.