data-analysis
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data (CSV, JSON, SQL), creating a surface for indirect prompt injection if those files contain malicious instructions.
- Ingestion points: Use of
pd.read_csv('data.csv')and SQLSELECTqueries inSKILL.md. - Boundary markers: Absent; the skill does not use delimiters to distinguish data from instructions.
- Capability inventory: The skill is allowed access to the
Bashtool and performs file system operations (writing PNG files) via Python. - Sanitization: Absent; data content is processed without prior validation or sanitization.
- [SAFE]: References to external documentation point to well-known and trusted official domains including pandas.pydata.org, matplotlib.org, and seaborn.pydata.org.
Audit Metadata