shopware-ddev
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFECREDENTIALS_UNSAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill facilitates the execution of arbitrary commands within a DDEV container via
ddev execandddev composer, which grants broad access to the application environment. - [CREDENTIALS_UNSAFE]: The documentation provides an example command for creating an administrative user that includes a hardcoded password string (
--password="secret"). While intended as a placeholder, it lacks explicit warnings to replace it with a secure value. - [INDIRECT_PROMPT_INJECTION]: The skill defines patterns for executing commands that could be exploited if untrusted external data is passed as arguments to the CLI tools.
- Ingestion points: Command templates in SKILL.md that take variable inputs.
- Boundary markers: None identified; commands are presented as direct execution strings without delimiting markers for external data.
- Capability inventory: The skill covers file system operations (database export/import), network configuration (port exposure), and administrative account management.
- Sanitization: No evidence of input validation or sanitization is present in the instruction set.
Audit Metadata