pricing-evaluation
Pass
Audited by Gen Agent Trust Hub on Aug 24, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external sales reports and event data through tools like
analysis report-data runandanalysis adhoc run. • Ingestion points: The skill pulls data from external reports and ad-hoc queries as part of the evaluation workflow in SKILL.md. • Boundary markers: Instructions do not contain explicit delimiters or warnings to separate external data from system instructions. • Capability inventory: The agent has tools to execute queries, create reports, and generate dashboard assets. • Sanitization: No data sanitization or validation logic is specified for the ingested content. - [COMMAND_EXECUTION]: The skill is designed to interact with the
ae-clitoolset to perform its primary analysis functions. This includes constructing JSON-based query definitions for event analysis and executing management commands for platform resources like dashboards and reports.
Audit Metadata