solutions-architect
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill uses local Python scripts to process architecture and workload data from external CSV and JSON files. This creates an attack surface where malicious instructions could be embedded in the data to influence the agent's subsequent analysis or documentation output.
- Ingestion points: Data is loaded via
scripts/architecture_scorer.py,scripts/migration_assessor.py, andscripts/sizing_calculator.pyusing the--dataargument. - Boundary markers: The scripts do not implement specific delimiters or 'ignore' instructions for content within the data fields.
- Capability inventory: The scripts are restricted to local file reading and console output, with no network access, file writing, or dynamic execution capabilities, limiting the potential impact of an injection.
- Sanitization: The scripts utilize standard library parsers (json, csv) but do not perform validation or sanitization of the content for natural language instructions.
Audit Metadata