csv-data-summarizer
Pass
Audited by Gen Agent Trust Hub on May 12, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill instructions contain forceful directives designed to override standard agent behavior. Specifically, the use of phrases like 'CRITICAL BEHAVIOR REQUIREMENT', 'DO NOT ASK THE USER WHAT THEY WANT', and 'IMMEDIATELY AND AUTOMATICALLY' attempts to bypass the agent's interactive safety protocols and force immediate execution without user confirmation.
- [PROMPT_INJECTION]: The skill exhibits vulnerability to indirect prompt injection due to its processing of untrusted data.
- Ingestion points: Untrusted data enters the agent context through the analysis of user-provided CSV files via the
summarize_csvfunction. - Boundary markers: The instructions lack delimiters or 'ignore embedded instructions' warnings to separate the operational instructions from the external data content.
- Capability inventory: The skill possesses capabilities for Python code execution, data manipulation with pandas, and visualization generation with matplotlib and seaborn.
- Sanitization: The skill description does not mention any sanitization, validation, or escaping of the content extracted from processed files.
- [COMMAND_EXECUTION]: The skill utilizes Python-based tools to perform data analysis. Because the instructions explicitly command the agent to suppress user interaction and clarification, malicious content hidden within a CSV file could potentially influence agent actions or trigger unintended processing without human oversight.
Audit Metadata