bulk
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes lists of items and passes them into sub-agent threads as prompts.
- Ingestion points: Item lists are accepted via user invocation or skill delegation in SKILL.md.
- Boundary markers: Instructions explicitly require the agent to keep item text separate from instructions and prevent item content from overriding the common prompt.
- Capability inventory: The skill utilizes thread management tools including create_thread, send_message_to_thread, and set_thread_archived.
- Sanitization: The skill relies on natural language boundary enforcement and strict prompt construction to mitigate injection risks.
- [COMMAND_EXECUTION]: The skill orchestrates workflows using authorized platform tools such as list_projects and list_threads to resolve environment context and manage sub-tasks.
Audit Metadata