ClawHub
Back to skill

Security audit

Meeting Prep

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent meeting-prep helper, but it may research people and pull prior notes or CRM data without clear confirmation or limits.

Review before installing if your agent has access to private notes, CRM systems, or client history. Use it only where attendee research and relationship-history summaries are appropriate, and prefer confirming each run before it uses internal records.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Vague Triggers

Medium
Confidence
96% confidence
Finding
The trigger phrases are broad enough that ordinary requests like 'brief me on' or 'I have a meeting with' could activate the skill without clearly signaling that the agent will research people online and may pull prior notes or CRM data. This creates a consent and privacy risk because the user may not realize the skill will collect, aggregate, and surface potentially sensitive information about attendees and past interactions.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill instructs the agent to research individual attendees, including LinkedIn activity and public statements, but provides no warning that personal data will be gathered and summarized. In context, this is more dangerous because the feature is explicitly designed to profile named individuals for a real-world meeting, increasing the chance of unnoticed collection and use of personal information.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill tells the agent to pull previous notes or CRM data if the user has met the person or company before, but it does not warn that internal records, prior conversations, or potentially confidential relationship history may be surfaced in the briefing. This is especially risky because CRM and meeting notes often contain sensitive business data, personal details, and commitments that could be exposed to the wrong user or included without need-to-know review.

Static analysis

No suspicious patterns detected.