ClawHub
Back to skill

Security audit

Pans Meeting Summarizer

Security checks across malware telemetry and agentic risk

Overview

This skill is a straightforward meeting summarizer, but users should know meeting transcripts may be sent to OpenAI and the advertised CRM sync is not actually implemented.

Install only if your meeting audio and transcripts are permitted to be processed by OpenAI. Use a dedicated OpenAI API key with spending controls, avoid feeding regulated or highly confidential meetings without review, and treat CRM sync as non-functional unless a future version adds and documents a real CRM integration.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Lp3

Medium
Category
MCP Least Privilege
Confidence
84% confidence
Finding
The skill advertises operations that imply access to local files and likely environment-backed credentials, but it does not declare any permissions or boundaries for those capabilities. This creates a transparency and governance gap: users and hosting platforms may not understand that meeting transcripts, audio files, output notes, and CRM credentials could be accessed or modified.

Vague Triggers

Medium
Confidence
79% confidence
Finding
The trigger phrases are very broad and generic, including common terms like '会议总结', 'meeting minutes', and 'action items', which increases the chance of accidental invocation in unrelated contexts. Because the skill can process sensitive meeting content and potentially sync results to a CRM, unintended activation could expose private business information or cause unwanted downstream actions.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill mentions one-click CRM synchronization but does not clearly warn that this can modify external systems and customer records. In a sales context, meeting notes often contain sensitive commercial data, so silent or poorly disclosed syncing raises the risk of unintended data propagation, privacy issues, and inaccurate CRM updates.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The code sends full meeting transcript content to the OpenAI API, which may include sensitive business discussions, customer data, and personal information, but it provides no explicit consent flow, privacy warning, or data-handling disclosure to the user. In the context of sales meeting summarization and CRM workflows, this is especially risky because transcripts are likely to contain confidential commercial information and regulated personal data.

VirusTotal

67/67 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.