press-media-relations
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill contains no executable code, hidden commands, or malicious patterns. It is comprised of markdown instructions and configuration schemas designed for marketing workflows.
- [INDIRECT_PROMPT_INJECTION]: The instructions in
SKILL.mdand theUnified Skill Contractexplicitly mandate that external data such as scraped web content, emails, and third-party exports must be treated as untrusted evidence rather than instructions, mitigating potential prompt injection attacks from external sources. - [CREDENTIALS_UNSAFE]: The skill includes security guidelines in
references/connectors.mdthat prohibit hardcoding API keys and sensitive credentials within the skill's logic or version control system. - [COMMAND_EXECUTION]: The skill defines clear boundaries for actions, requiring explicit user authorization for any persistent writes or external side effects, ensuring the agent operates within a controlled and approved scope.
Audit Metadata