sustainability-advisor
Pass
Audited by Gen Agent Trust Hub on Jul 31, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists entirely of Markdown and JSON files providing instructions and metadata. No executable scripts, network requests, or sensitive file accesses were detected.
- [PROMPT_INJECTION]: The instructions contain natural phrases such as 'Never do this' and 'Do not use it for,' which are benign constraints intended to define the skill's operational boundaries and prevent the agent from providing legal advice or glossy marketing copy. No malicious override or jailbreak patterns were found.
- [DATA_EXFILTRATION]: There are no commands that attempt to access local credentials, environmental variables, or perform unauthorized network operations. All data processing is confined to the user-provided context.
- [REMOTE_CODE_EXECUTION]: The skill does not include or reference any external scripts, package installations (npm/pip), or remote commands.
- [INDIRECT_PROMPT_INJECTION]: While the skill processes user-supplied business data (such as emissions hotspots and company goals), it includes explicit instructions for the AI to maintain boundary clarity, state what is unknown, and prioritize evidence-backed conclusions, which serves as a defensive mechanism against being misled by untrusted data.
Audit Metadata