php-best-practices
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions in
SKILL.mdandAGENTS.mddirect the agent to execute shell commands (php -v,grep) to verify the local PHP runtime version and project dependencies incomposer.json. - [INDIRECT_PROMPT_INJECTION]: The skill acts on external, untrusted data by performing code reviews on project files.
- Ingestion points: The agent reads PHP source code and
composer.jsonfrom the local file system (identified inSKILL.md). - Boundary markers: The instructions lack explicit delimitation or instructions for the agent to ignore prompt-like content within the analyzed files.
- Capability inventory: The agent can read files and execute shell commands to query the environment.
- Sanitization: No specific sanitization or filtering of the code content is performed before processing.
Audit Metadata