specialist-design-whimsy-injector
Pass
Audited by Gen Agent Trust Hub on Aug 15, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user data and interpolates it into a subagent prompt without using boundary markers or sanitization.
- Ingestion points: The
{the user's question}variable in the Agent tool call withinSKILL.mdcaptures arbitrary user input. - Boundary markers: Absent. The user input is concatenated directly after the string 'Answer: ', providing no separation between the system's instruction and the user's content.
- Capability inventory: The skill invokes the
Agenttool to spawn subagents, but does not perform file writes, network operations, or command execution itself. - Sanitization: No evidence of validation, escaping, or filtering of the user-provided content before it is processed by the agent.
Audit Metadata