specialist-engineering-sre
Pass
Audited by Gen Agent Trust Hub on Aug 15, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by interpolating raw user input into a sub-agent's system prompt.
- Ingestion points: The variable
{the user's question}inSKILL.mdaccepts untrusted external data. - Boundary markers: No delimiters, such as triple quotes or XML tags, or explicit 'ignore embedded instructions' warnings are used to isolate the user input from the sub-agent's instructions.
- Capability inventory: The sub-agent is assigned a broad persona with authority over production reliability decisions, which could be influenced by malicious content in a user question.
- Sanitization: The skill does not perform any validation, filtering, or escaping of the user input before interpolation.
Audit Metadata