marketing-plan
Pass
Audited by Gen Agent Trust Hub on Nov 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill has a defined attack surface as it ingests untrusted data from external sources such as web scrapes, reviews, and emails. However, it implements robust defenses by explicitly instructing the agent to treat all such data as evidence rather than instructions, and specifically mandates that the agent must not follow any instructions embedded within that data.
- [DATA_EXFILTRATION]: The skill contract strictly limits side effects. It prohibits publishing, sending data, or mutating accounts without explicit user authorization. It also specifies that recommendations should be written to the response by default rather than persistent files.
- [CREDENTIALS_UNSAFE]: The reference documentation follows security best practices by explicitly instructing developers to keep API keys out of the skill source code and configuration files.
- [PATH_TRAVERSAL]: The instructions include a specific safety check for client artifact creation, requiring the agent to validate identifiers and reject path traversal sequences (separators) to ensure file operations remain within the intended directory.
Audit Metadata