rfm-customer-segmentation

Pass

Audited by Gen Agent Trust Hub on Mar 12, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill ingests and processes user-supplied CSV files, which serves as a potential vector for indirect prompt injection. Malicious instructions embedded in the data could attempt to manipulate the agent's summaries or subsequent actions.
  • Ingestion points: External transaction data is loaded using pd.read_csv in core_analysis.py.
  • Boundary markers: The skill instructions in SKILL.md do not include explicit delimiters or safety prompts to ignore potential natural language instructions within the processed data.
  • Capability inventory: The skill is granted Bash execution and Write permissions to perform analysis and generate report files.
  • Sanitization: No explicit sanitization or filtering of the content within the CSV data is performed before the agent interprets the results.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 12, 2026, 05:34 AM