review-offered-task
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to retrieve task information from an external source, creating a potential surface for indirect prompt injection if the task data contains adversarial content.
- Ingestion points: The
get-task-detailstool inSKILL.mdfetches external data that the agent is then directed to evaluate. - Boundary markers: No specific delimiters or safety warnings are provided to the agent for handling the content returned by
get-task-details. - Capability inventory: The skill uses
task-actionfor state transitions and invokes thework-on-taskskill for execution. - Sanitization: There are no instructions for the agent to sanitize or validate the content of the task description before processing.
Audit Metadata