messenger-retry-failures
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides instructions for the agent to manage and inspect asynchronous message queues using shell commands such as
bin/console messenger:failed:show. These queues typically contain data from external sources, such as payment gateway responses or user-submitted order details, creating a surface where malicious payloads could be ingested into the agent's context. - Ingestion points: Failed message payloads and stack traces retrieved via the terminal in
reference.md. - Boundary markers: No delimiters or warnings are provided to help the agent distinguish between command output data and instructions.
- Capability inventory: The skill environment allows for
shell(Bash) andeditcapabilities as specified inSKILL.md. - Sanitization: The implementation patterns do not include sanitization for external error messages or data payloads before they are displayed in the terminal or processed in notifications.
Audit Metadata