twilio-outbound-call
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources including PagerDuty, pyATS, and NetBox to generate messages for outbound Twilio calls. This creates a surface where instructions embedded in external status data could influence the agent's voice output.
- Ingestion points: The skill queries multiple MCP servers (PagerDuty, pyATS, NetBox) in both the On-Demand and Scheduled workflows defined in SKILL.md.
- Boundary markers: While the skill includes instructions to sanitize content (removing IPs and credentials), it lacks explicit delimiters or instructions to ignore embedded commands within the ingested data.
- Capability inventory: The skill utilizes the
twilio_voice_calltool to perform outbound communication. - Sanitization: Sanitization is focused on preventing sensitive data exposure (IPs/credentials) but does not address input validation for the generated message content.
Audit Metadata