meeting-to-tasks
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted external data in the form of meeting transcripts which may contain malicious instructions.
- Ingestion points: Meeting transcripts provided via plain text, Markdown, or audio transcript exports (Otter.ai, Fireflies, Zoom, Google Meet, Microsoft Teams) as described in the workflow of SKILL.md.
- Boundary markers: The instructions do not define clear delimiters or specific instructions for the agent to ignore potentially malicious commands embedded within the transcript content during extraction.
- Capability inventory: The skill allows the use of Read, Write, Bash, and Grep tools, providing a capability surface that could be targeted if an agent is tricked into executing commands found within the transcript data.
- Sanitization: While the skill guidelines suggest flagging sensitive HR or confidential information, there is no technical sanitization or escaping of extracted text before it is inserted into the task templates for Linear, GitHub, or other tools.
Audit Metadata