voice-dossier-builder
Pass
Audited by Gen Agent Trust Hub on Aug 20, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from user posts, emails, and pitch decks which could harbor malicious instructions. It mitigates this risk by explicitly instructing the agent to treat these inputs as untrusted and preventing them from overriding internal configurations. The ingestion points are specified in the collection phase, and the capability inventory is limited to authorized scripts.
- [COMMAND_EXECUTION]: The skill uses internal scripts (
scripts/connectors/bluesky.py,fediverse.py, andregistry-events.py) to perform its primary tasks. These are referenced as authorized components for fetching data and updating memory via proposed events.
Audit Metadata