ClawHub
Back to skill

Security audit

AgentMeet

Security checks for vulnerabilities and agentic risk

Overview

AgentMeet is a coherent scheduling skill, but it asks for Gmail and Google Calendar authority while leaving important approval, validation, and install-code boundaries unclear.

Review before installing. Only use this if you can verify the actual AgentMeet source code and dependencies at the referenced path, grant the smallest possible Google permissions, and require the agent to show each outgoing email, parsed reply, and calendar event before sending or creating anything.

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 (2)

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill documentation explicitly says it requires Google Calendar and Gmail access and sends protocol-formatted meeting emails, but it does not present a prominent user warning about the sensitivity of calendar contents and meeting metadata being accessed and transmitted. This can lead users to authorize broad access without understanding that availability patterns, participant identities, and meeting details may be exposed through outbound email workflows.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The workflow states that the sender's agent detects replies and creates calendar events for both parties, but it does not mention any confirmation gate, validation step, or user approval before taking that action. In practice, this could cause unauthorized or mistaken event creation from malformed, spoofed, or misinterpreted email replies, especially in a protocol that operates over plain email.

Static analysis

No suspicious patterns detected.