meta-strategic-factor-analysis
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from untrusted external sources, including outputs from a Digital research engine, client records, and data provided by other skills in a multi-step orchestration pipeline. These ingestion points create a surface for indirect prompt injection where instructions hidden in the processed research data could influence the agent's logic during the weighted factor analysis.\n
- Ingestion points: External evidence from the Digital research engine, client files (specifically referenced as 02, 03, 08, 09, 10), and country-context/ files.\n
- Boundary markers: The instructions do not specify the use of delimiters or 'ignore embedded instructions' warnings when processing retrieved evidence to prevent the agent from accidentally executing commands found in the source material.\n
- Capability inventory: The skill is permitted to read and search evidence files and draft complex strategic tables (EFAS, IFAS, SFAS, TOWS) based on the input.\n
- Sanitization: There are no explicit instructions for sanitizing or validating the content of external research data or client records before they are interpolated into the strategic analysis workflow.
Audit Metadata