brainstorming
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources during the exploration and understanding phases, which introduces an inherent surface for indirect prompt injection if the retrieved content contains instructions designed to influence the agent.
- Ingestion points: Findings from the
@agent-ed3d-research-agents:internet-researchersub-agent and results fromWebSearchandWebFetchoperations. - Boundary markers: The skill does not explicitly instruct the agent to use delimiters or ignore embedded instructions when processing external research content.
- Capability inventory: The skill's active capabilities are limited to workflow management tools (
TaskCreate,TaskUpdate) and user interaction (AskUserQuestion). It does not have direct code execution or file-write permissions. - Sanitization: No explicit content sanitization or validation protocols are defined for the data retrieved from external sources.
- [EXTERNAL_DOWNLOADS]: The skill references external tools and agents to fulfill its research requirements.
- It delegates tasks to
@agent-ed3d-research-agents:internet-researcher, which is a specialized agent associated with the skill's authoring vendor,ed3dai. - It utilizes standard
WebSearchandWebFetchtools to access external API documentation and library information.
Audit Metadata