retention-metric-to-event-report
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
ae-clicommand-line tool to perform operations on the analytics platform, including creating, updating, and deleting metrics and reports. It constructs shell commands by interpolating JSON definitions into single-quoted strings, which presents a potential command injection surface if the platform data contains unescaped shell-special characters.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from an external platform and uses it to perform consequential actions, creating a vulnerability surface.\n - Ingestion points: The agent retrieves project metadata, report definitions, and metric configurations via
ae-cli analysis report getandae-cli analysis-meta metric getcommands.\n - Boundary markers: The instructions lack specific guidance for the agent to use delimiters or ignore instructions that might be embedded in the retrieved data.\n
- Capability inventory: The agent has broad write and delete capabilities on the platform using the
ae-clitool.\n - Sanitization: The skill does not define procedures for sanitizing or validating the retrieved data before it is reused in new commands.
Audit Metadata