pmstudio-meeting-notes
Pass
Audited by Gen Agent Trust Hub on Nov 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of meeting transcripts, creating a surface for indirect prompt injection where malicious instructions embedded in a transcript could influence the agent's behavior.
- Ingestion points: Raw meeting transcripts provided via chat, file paths, or audio transcription output as described in Step 1 and Step 3.
- Boundary markers: The instructions lack explicit delimiters or boundary markers to isolate the untrusted transcript content from the system instructions.
- Capability inventory: The skill is authorized to read project-specific files such as
CLAUDE.local.md, Stakeholder directories (Excel and Markdown), and PRDs. It also has the capability to write new Markdown files to project-specific subfolders. - Sanitization: There are no documented procedures for sanitizing, filtering, or validating the ingested transcript content before processing.
Audit Metadata