php
Pass
Audited by Gen Agent Trust Hub on Jul 3, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes untrusted PHP source code, composer.json, and CI configurations, which presents an attack surface for indirect prompt injection. Malicious instructions hidden in the audited code could attempt to influence the agent's behavior.
- Ingestion points: Reads PHP source files, CI matrices, and composer.json files from the repository.
- Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions when reading external file content.
- Capability inventory: Uses shell commands like grep and interacts with static analysis tools such as PHPStan and Rector.
- Sanitization: The skill recommends corroborating hits and using AST-aware tools to filter comments and strings, but does not specify input sanitization for the agent's prompt context.
Audit Metadata