dd-aws-integration
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses
terraform,awsCLI, andcurlto provision AWS resources and interact with the Datadog API. These actions are within the scope of the skill's purpose and the tool definitions. - [DATA_EXFILTRATION]: The skill reads credentials from
.envand.env.localfiles and transmits them to official Datadog API endpoints (api.datadoghq.comand regional variants) for validation and resource management. This is consistent with the vendor's own infrastructure and standard authentication patterns for integration tools. - [INDIRECT_PROMPT_INJECTION]: The skill processes data from several external sources that could theoretically be manipulated to influence agent behavior.
- Ingestion points: Reads environment files (
.env,.env.local), local Terraform configuration files (.tf,terraform.tfstate), AWS CLI command output, and JSON responses from the Datadog API. - Boundary markers: None identified; data is processed directly as shell variables or CLI inputs.
- Capability inventory: The skill has broad capabilities including shell execution, file system access, and infrastructure management via Terraform.
- Sanitization: None identified; however, the skill uses structured HCL templates and standard shell patterns which mitigate common injection risks.
Audit Metadata