support-engineer
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted data from external sources, such as customer tickets, error messages, and server logs. This presents a potential surface for indirect prompt injection if malicious instructions are embedded in the data provided by customers.
- Ingestion points: Customer-reported bugs, API error payloads, and server logs (
SKILL.md,references/reproduction_debugging.md). - Boundary markers: The templates do not use specific delimiters or instructions to prevent the agent from following embedded instructions in the ingested data.
- Capability inventory: The skill involves debugging tasks and reproduction steps (e.g., using
curl), which could be influenced by injected instructions. - Sanitization: The skill includes best practices for data handling, specifically instructing to redact secrets and truncate PII in
references/reproduction_debugging.md.
Audit Metadata