order-types-instructions
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted data from external CSV files, which presents a surface for indirect prompt injection if the resulting diagnostics are subsequently interpreted as instructions by the AI agent.
- Ingestion points: The script
scripts/order_types_instructions_diagnostics.pyreads user-provided CSV files viapandas.read_csv. - Boundary markers: There are no explicit instructions or delimiters provided to the agent to distinguish between the data being processed and instructions.
- Capability inventory: The skill has the capability to read from the local file system and write JSON output files.
- Sanitization: The diagnostic script performs basic data validation by attempting to coerce input columns into numeric and datetime formats, which provides a minimal layer of protection against non-conforming data types.
Audit Metadata