ec2
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEOBFUSCATIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [OBFUSCATION]: The file 'instance-management.md' includes a Base64-encoded string within a JSON payload for a launch template creation command. While this is a standard requirement for the AWS CLI UserData parameter, it is technically obfuscated executable code.
- Evidence: '"UserData": "IyEvYmluL2Jhc2gKeXVtIHVwZGF0ZSAteQo="' in 'instance-management.md'.
- Decoded content: '#!/bin/bash\nyum update -y\n'.
- [DYNAMIC_EXECUTION]: The skill demonstrates generating and executing scripts on remote instances through the 'UserData' field in launch configurations and commands across 'SKILL.md' and 'instance-management.md'.
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data from AWS CLI outputs, creating a vulnerability surface for indirect prompt injection.
- Ingestion points: Untrusted data enters the agent context through the output of commands like 'aws ec2 describe-instances', 'aws cloudwatch get-metric-statistics', and 'aws ssm start-session' as documented in 'SKILL.md' and 'instance-management.md'.
- Boundary markers: There are no explicit delimiters or instructions provided to the agent to disregard potential natural language instructions embedded within the command outputs.
- Capability inventory: The skill provides powerful capabilities including resource termination, security group modification, and establishment of remote interactive sessions (SSM/SSH).
- Sanitization: The skill lacks logic for sanitizing or validating the content of external tool outputs before processing them.
Audit Metadata