personize-signal
Pass
Audited by Gen Agent Trust Hub on Mar 13, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches functionality from official @personize/sdk and @personize/signal packages and references the personize.ai domain.
- [DATA_EXFILTRATION]: Properly manages API keys and webhook secrets using environment variables (PERSONIZE_SECRET_KEY, SLACK_WEBHOOK, WEBHOOK_SECRET).
- [PROMPT_INJECTION]: The skill architecture processes external event data alongside natural language governance rules, representing a surface for indirect prompt injection. Ingestion points: Event payloads in signal.trigger() and WebhookSource data. Boundary markers: Uses governance variables to guide AI decisions, though the snippets do not show explicit data delimiters. Capability inventory: Capable of sending notifications via Slack, Email, and SMS channels and modifying workspace memory. Sanitization: Relies on the internal engine of the @personize/signal package; no explicit data sanitization is shown in the integration code.
Audit Metadata