experimentation-and-ab-testing
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to read external data from other profiles such as 'brand-profile' and 'goals-and-kpis' to inform its experiment designs. This creates an attack surface where malicious instructions embedded in those profiles could influence the agent's output.
- Ingestion points: Reads data from
brand-profileandgoals-and-kpisin the initial design phase. - Boundary markers: No explicit delimiters or instructions to ignore embedded prompts within the ingested data are defined.
- Capability inventory: Generates content variants (hooks, captions, scripts) and interfaces with the
WoopSocialtool for scheduling. - Sanitization: No evidence of input validation or sanitization is present in the skill's instructions.
- [NO_CODE]: The skill consists entirely of Markdown instructions and reference files. It does not contain any Python scripts, Node.js code, or shell commands.
Audit Metadata