data-refresh-eval
Pass
Audited by Gen Agent Trust Hub on Feb 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the Bash tool to execute local CLI commands for dataset management and evaluation, specifically running
bun src/index.tswithin the vendor's repository. - [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection because it processes untrusted customer message data retrieved from the Front API during the evaluation process.
- Ingestion points: Untrusted conversation content is fetched from the Front API and stored in local files such as
data/eval-dataset.json. - Boundary markers: No boundary markers or specific instructions to disregard embedded commands were found in the workflow documentation.
- Capability inventory: The skill leverages
Bash(skill:*)for script execution and data manipulation using tools likecatandjq. - Sanitization: There is no evidence of sanitization or escaping of external content before it is analyzed by the agent response quality workflow.
Audit Metadata