budget-recommendation-calculator
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The README.md provides a command to fetch the skill's instruction file from the author's official GitHub repository. This uses a well-known service (GitHub) and targets the vendor's own resource.
- [INDIRECT_PROMPT_INJECTION]: The skill processes performance metrics from external sources such as spreadsheets or API query results.
- Ingestion points: MTD spend, target budgets, and impression share data (README.md, SKILL.md).
- Boundary markers: None defined in the instructions.
- Capability inventory: The skill's execution environment is restricted to the
Readtool via the frontmatter, preventing the agent from performing unauthorized file writes, network requests, or command execution based on untrusted data. - Sanitization: Not present, but the risk is negligible due to the restricted toolset.
- [SAFE]: The skill implements disciplined budget logic, including a mandatory 10% cap on increases and a 7-day ramp-up period, ensuring safe and predictable behavior.
Audit Metadata