Security audit
Security checks for vulnerabilities and agentic risk
This meeting-summary skill is a plain Markdown instruction file that processes user-provided meeting content and does not install code, access tools, or persist data on its own.
Use this skill for meeting transcripts or notes you are comfortable sharing with your agent. Avoid including confidential, regulated, or highly personal details unless your organization permits it, and be mindful that broad triggers like action items or follow-up email may activate it for meeting-adjacent requests.
The manifest trigger list includes generic phrases like "action items," "follow-up email," and "meeting notes" without contextual constraints. These are common workplace requests and could cause unintended invocation when the user is not explicitly asking for this skill.
The activation section says the skill activates for broad intents like "meeting summarization" and "follow-up email drafting," but does not clarify required inputs, context, or when similar requests should not trigger the skill. Without negative examples or tighter constraints, the activation boundary remains ambiguous.
This skill processes meeting transcripts, attendee identities, decisions, blockers, and follow-up content, all of which commonly contain sensitive business or personal information. Omitting a warning or handling guidance increases the chance that users will submit confidential material without awareness of privacy, retention, or sharing risks.
The statement that the skill can maintain context across sessions if history is shared introduces a persistence expectation for potentially sensitive meeting content. Without explicit retention limits, consent requirements, access boundaries, and minimization rules, recurring use could cause confidential discussions, participant data, or prior decisions to bleed into later sessions or be retained longer than intended.
- Works with any transcript format (Zoom, Google Meet, Teams, or raw text)
- Can accept timestamps — will annotate key moments
- Best results with clear speaker labels, but handles raw text too
- For recurring meetings: can maintain context across sessions if history is shared
---
No suspicious patterns detected.