well-actually
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection by processing external, untrusted content.
- Ingestion points: The sub-agents ingest data from rendered websites, DOM structures, and project source code (referenced in
SKILL.mdunder 'Code Access Rules'). - Boundary markers: The 'Agent Prompt Template' does not include delimiters or specific instructions to disregard commands that might be embedded in the text, HTML, or code comments being reviewed.
- Capability inventory: The agent is granted capabilities to browse the web (via Browser MCP) and read local source code files.
- Sanitization: There are no defined processes for sanitizing or escaping the content retrieved from external URLs or local files before it is processed by the persona-based sub-agents.
Audit Metadata