aws-security-scanner
Pass
Audited by Gen Agent Trust Hub on Aug 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill's primary purpose is to audit AWS security, which requires accessing sensitive configurations, IAM policies, and credential metadata. This capability creates an exposure surface for sensitive account data.
- [COMMAND_EXECUTION]: The skill utilizes the
Bashtool to run external security scanners such as Prowler. This provides a broad execution capability within the agent environment. - [PROMPT_INJECTION]: The skill ingests untrusted data from the AWS environment (e.g., resource tags, bucket names, and policy descriptions). This introduces a risk of indirect prompt injection where malicious instructions embedded in the audited infrastructure could attempt to influence the agent's reasoning or output.
- Ingestion points: AWS resource configurations and metadata retrieved during scanning (SKILL.md).
- Boundary markers: None identified in the skill instructions to separate audited data from agent instructions.
- Capability inventory: Bash, Read, Write, Glob, Grep, WebFetch (SKILL.md frontmatter).
- Sanitization: No explicit sanitization or validation of the audited resource data is described.
Audit Metadata