preference-frequency-manager
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE]: The skill is composed entirely of Markdown documentation and reference files. It contains no scripts (.py, .js, .sh), binary executables, or dynamic code execution blocks.
- [SAFE]: No malicious patterns, obfuscation, or suspicious network activities were identified during the analysis.
- [INDIRECT_PROMPT_INJECTION]: The instructions in
SKILL.mdand theUnified Skill Contractinreferences/skill-contract.mdspecifically mandate treating external inputs (scraped text, reviews, comments) as untrusted evidence. This is a proactive safety measure to prevent the agent from following instructions embedded in third-party data. - [CREDENTIALS_UNSAFE]: The
references/connectors.mdfile explicitly states that API keys must be kept out of skills and source control, demonstrating alignment with secure credential management standards. - [PRIVILEGE_ESCALATION]: The skill framework requires explicit user authorization for any persistent writes, real-world side effects, or account mutations, ensuring the agent operates under the principle of least privilege.
Audit Metadata