influencer-brief
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes content from local brand profile and guideline files located in the
~/.claude-marketing/directory. This creates an attack surface where instructions embedded within these files could influence the agent's behavior during the brief generation process. - Ingestion points: The skill reads data from
~/.claude-marketing/brands/{slug}/profile.json,~/.claude-marketing/brands/{slug}/guidelines/_manifest.json, custom templates, and agency SOPs. - Boundary markers: The instructions do not specify the use of delimiters or specific system-level warnings to ignore natural language instructions that might be contained within the loaded brand data.
- Capability inventory: The skill is primarily a text generation tool and does not appear to have access to dangerous system commands, file writing, or network operations.
- Sanitization: There is no evidence of input validation or sanitization for the JSON and Markdown files loaded from the brand directories.
Audit Metadata