budget-reallocation
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill incorporates comprehensive security guidelines, specifically instructing the agent to treat fetched or pasted content (such as campaign names, labels, or website data) as data to be reported on, never as instructions to be followed.
- [SAFE]: A strict read-only policy is enforced. The skill is designed to propose budget reallocation scenarios for human approval but explicitly states it "moves nothing" and never creates rules or executes transfers directly.
- [SAFE]: The skill follows clear data privacy and security protocols, including instructions to never echo or persist credentials, redact tokens from output, and avoid following URLs found within untrusted content.
- [SAFE]: Interaction constraints, such as the three-question limit and the use of stated assumptions, are formatted for user efficiency and autonomy. The skill requires the agent to be transparent about what information was inferred rather than observed, ensuring the user remains informed of the reasoning process.
- [SAFE]: Data integrity checks are mandatory before analysis, ensuring that normalization (currency, timezones, attribution windows) occurs and that volume floors are applied to prevent drawing conclusions from statistically insignificant noise.
Audit Metadata