voice-memo-triage
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes shell utilities including
cp,grep,sed, andwcto manage file backups, identify content boundaries, and perform line-based deletions. It also executesgit committo persist changes in a local vault. - [INDIRECT_PROMPT_INJECTION]: The skill ingests and triages untrusted text from a voice memo file, which is then used to formulate research directions for sub-agents in a workflow.
- Ingestion points: File content is read from the aggregate markdown file specified in the
aggregateparameter. - Boundary markers: Entries are delimited by Markdown headers matching the weekday/date/time pattern.
- Capability inventory: The skill possesses the ability to read and write files, execute shell commands, and perform version control operations.
- Sanitization: No specific sanitization or filtering of the memo content is described before it is passed to workflow agents; the skill explicitly preserves the content "byte for byte".
Audit Metadata