awesome-marketing-science-guide
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions for installing marketing science libraries from established sources.
- Fetches R libraries such as Robyn and GeoLift from Meta's (Facebook) official GitHub repositories.
- Recommends standard Python packages including Google's LightweightMMM and Microsoft's EconML via public registries.
- [INDIRECT_PROMPT_INJECTION]: The skill describes workflows for ingesting and processing marketing data, which is a standard operational requirement for measurement tools.
- Ingestion points: Data is ingested from BigQuery tables and pandas DataFrames within the modeling functions (e.g.,
load_marketing_data_for_mmminSKILL.md). - Boundary markers: None explicitly defined within the code snippets.
- Capability inventory: Accesses BigQuery for data retrieval and performs statistical modeling using standard scientific libraries.
- Sanitization: Standard data science workflows described do not include explicit input sanitization, which is typical for this domain.
Audit Metadata