aws-solution-architect

Warn

Audited by Gen Agent Trust Hub on Feb 19, 2026

Risk Level: MEDIUMCOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION] (MEDIUM): The skill workflow executes multiple Python scripts (architecture_designer.py, serverless_stack.py, cost_optimizer.py) and references several documentation files in the references/ directory that are missing from the analyzed package. This prevents a full security audit of the actual code being executed.
  • [COMMAND_EXECUTION] (MEDIUM): The skill instructs the agent to run high-privilege AWS CLI commands, specifically aws cloudformation create-stack with the --capabilities CAPABILITY_IAM flag. This allows the skill to create or modify IAM roles, which can lead to privilege escalation if the generated templates are malicious or overly permissive.
  • [INDIRECT_PROMPT_INJECTION] (LOW): The skill processes untrusted user data from requirements.json to generate infrastructure designs, creating a potential surface for instruction injection. 1. Ingestion points: requirements.json read by the architecture_designer.py tool. 2. Boundary markers: Absent; no delimiters or instructions to disregard embedded commands in the input data are present. 3. Capability inventory: Execution of Python scripts, AWS CLI, CDK, and Terraform. 4. Sanitization: No validation or escaping of the input JSON fields is documented.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 19, 2026, 03:56 PM