aws-fis-experiment-execute
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands that directly interpolate user-supplied strings (such as the template ID input) or directory names into Bash snippets. For example, Step 1a uses a find command with the template ID variable, and Step 1b uses basename and grep on the experiment directory name. This creates a command injection vulnerability if the agent fails to sanitize these inputs before execution.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external experiment directories and application log streams to generate summaries and detect recovery signals. * Ingestion points: experiment README.md files and continuous kubectl log streams. * Boundary markers: The instructions do not define delimiters or specific warnings to distinguish processed data from the agent's instructions. * Capability inventory: The skill has the ability to execute AWS CLI commands (start/stop experiments), call other skills, and write results to the local file system. * Sanitization: No explicit sanitization or filtering logic is provided for the ingested logs or metadata.
Audit Metadata