cloud-security-engineer
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process and analyze external data sources, including Infrastructure-as-Code (IaC) templates (Terraform, CloudFormation, Bicep, Pulumi) and cloud environment configurations. This creates a surface for indirect prompt injection where an attacker could embed malicious instructions within the audited files to influence the agent's behavior.
- Ingestion points: Reads and analyzes external IaC files and cloud configurations provided in the workspace.
- Boundary markers: The skill includes a specific "Prompt Injection Safety Notice" section instructing the agent to ignore conflicting instructions found in processed data.
- Capability inventory: The skill is configured to use
Read,Grep, andGlobtools to inspect file contents. - Sanitization: Relies on natural language instructional guardrails within the prompt rather than programmatic filtering or strict schema validation of ingested content.
- [EXTERNAL_DOWNLOADS]: The skill contains reference links to security benchmarks and architecture frameworks from reputable organizations including the Center for Internet Security (CIS), NIST, AWS, Microsoft, and Google. These are informative documentation links used to support the posture review methodology and do not include executable code downloads.
Audit Metadata