Security audit

Meeting

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed meeting-assistant skill that drafts summaries and tasks for user review rather than secretly recording or automatically sharing content.

Before installing, make sure your team is comfortable with meeting notes being processed through your agent and host connectors. Review drafts before sharing, verify the target audience, and avoid adding confidential or regulated meeting content unless you have authorization.

SkillSpector

Vulnerability Patterns

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

Findings (3)

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The example normalizes capturing and synthesizing identifiable meeting content about named individuals, decisions, dissent, and task assignments without any warning about consent, privacy expectations, or handling of sensitive business information. In a meeting-assistant skill, this can encourage users to process and redistribute personal or confidential workplace data in contexts where participants may not expect recording or summarization.

Missing User Warnings

Medium

Confidence: 96% confidence
Finding: The example suggests syncing post-meeting outputs to a group channel, which implies external sharing of synthesized meeting content, but provides no warning, approval flow detail, or audience verification. This is risky because summaries often contain sensitive decisions, assignments, or dissent, and accidental dissemination to a broader group can cause confidentiality and privacy breaches.

Vague Triggers

Medium

Confidence: 89% confidence
Finding: The trigger phrases are generic operational phrases such as 'generate minutes' and 'track commitments' with no clear scoping to meeting-specific contexts, making accidental or overly broad invocation more likely. In an agent environment, this can cause the skill to activate in unrelated workflows and process sensitive notes, decisions, or commitments when the user did not explicitly intend to invoke it.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.