maintain-space
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes content from potentially untrusted external sources and user documents to update shared workspace content.
- Ingestion points: The skill reads from 'connected sources', 'source documents', 'page comments', and 'existing page text' as mentioned in SKILL.md.
- Boundary markers: The skill instructs the agent to 'Read the root and tool-identified instructions' and notes that 'source documents do not expand the user's authorization', providing logic-based boundaries.
- Capability inventory: The skill utilizes tools to write pages, create new pages, and schedule upkeep jobs.
- Sanitization: Instructions include 'Respect the destination's audience when using connected sources' and 'without exposing private source content', focusing on privacy and authorization rather than technical string sanitization.
Audit Metadata