specialist-engineering-php-engineer
Pass
Audited by Gen Agent Trust Hub on Aug 15, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill interpolates user-controlled data directly into the system prompt of a subagent.
- Ingestion points: The variable '{the user's question}' in SKILL.md.
- Boundary markers: Absent. The user input is appended at the end of the subagent prompt without delimiters or instructions to the model to treat the content as untrusted data.
- Capability inventory: The subagent is granted broad authority for PHP architecture, performance analysis, and security auditing.
- Sanitization: Absent. The input is not escaped or validated before interpolation.
Audit Metadata