data-analyst
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external datasets for statistical analysis and visualization. This creates a vulnerability surface where instructions embedded in the data (e.g., within CSV strings or metadata) could potentially influence the agent's behavior.
- Ingestion points: The instructions explicitly guide the agent to load and inspect datasets using commands like
df.head()anddf.describe()in SKILL.md. - Boundary markers: There are no explicit delimiters or instructions provided to the agent to disregard instructions potentially embedded within the datasets.
- Capability inventory: The skill assumes access to a Python execution environment with libraries capable of file manipulation and data processing (pandas, numpy).
- Sanitization: No sanitization or data validation steps specifically targeting indirect prompt injection are mentioned.
Audit Metadata