discover-brand
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from a wide variety of external, potentially untrusted sources which creates an attack surface for indirect prompt injection.
- Ingestion points: The skill searches and extracts content from Notion, Confluence, Google Drive, Box, Microsoft 365, Figma, Gong, Granola, and Slack.
- Boundary markers: There are no explicit delimiters or instructions provided to the sub-agent to distinguish between discovery data and instructions, making it vulnerable to embedded commands in the source files.
- Capability inventory: The skill utilizes the Task tool to launch a discovery agent and writes output to the local filesystem at
.claude/brand-voice-guidelines.md. - Sanitization: The instructions do not mention any sanitization, filtering, or validation steps for the content retrieved from external platforms before it is analyzed.
Audit Metadata