paper-ds-review
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill follows security best practices by explicitly instructing the agent to keep private file URLs and review artifacts out of public examples.
- [SAFE]: The workflow prevents unauthorized changes by stating that fixes and approvals should be handled by separate tools and not as a side effect of the review process.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from design comments and user edits. However, this is inherent to its primary purpose as a review tool, and the instructions emphasize recording evidence rather than executing commands based on that input.
- Ingestion points: Processes native design comments and user edits during the review workflow.
- Boundary markers: No specific delimiters or 'ignore' instructions are defined for processing external comments.
- Capability inventory: References the ability to load 'paper-ds-canvas' and mentions downstream tools for subsequent actions.
- Sanitization: No explicit sanitization of comment content is mentioned.
Audit Metadata