investigate-public-exposure
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill instructions and associated reference files provide a structured methodology for cloud security analysis. It focuses on identifying misconfigurations like open security groups, public S3 policies, and internet-facing load balancers.
- [DYNAMIC_EXECUTION]: The skill facilitates the generation of Cypher queries to interact with a security graph. It includes specific safety guidelines to prevent command injection, such as advising the use of parameters over string interpolation and safe quoting of literals when building queries.
- [INDIRECT_PROMPT_INJECTION]: The skill processes data retrieved from external cloud environments and security modules. It mitigates potential injection risks by requiring the agent to validate schemas (using
subimageGetNodesSchema) and verify findings against raw evidence before presenting them to the user. - [COMMAND_EXECUTION]: The skill uses specific MCP tools (
subimage-mcp) for database querying and resource lookup. These tools are scoped to the primary purpose of the skill and do not involve arbitrary shell command execution on the host system.
Audit Metadata