support-feedback-prioritization
Pass
Audited by Gen Agent Trust Hub on Apr 2, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill processes untrusted external data from Intercom tickets and Slack messages, which presents a surface for indirect prompt injection.
- Ingestion points: Step 1 specifies gathering all tickets and conversations from Intercom and messages from defined Slack channels.
- Boundary markers: The prompt does not utilize explicit delimiters (e.g., XML tags or triple quotes) to isolate user-generated feedback from the agent's core instructions.
- Capability inventory: The skill utilizes Intercom (read), Slack (read), CRM (read), and Google Drive (write) capabilities.
- Sanitization: There is no explicit instruction to sanitize or escape content found in the raw text of support signals.
- Mitigation: The skill includes a 'Read only' rule prohibiting the agent from sending messages or replying to tickets, which prevents common injection outcomes like automated phishing or unauthorized communication.
Audit Metadata