gcp-to-aws

Pass

Audited by Gen Agent Trust Hub on Nov 19, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONDYNAMIC_CONTEXT_INJECTION
Full Analysis
  • External Downloads: The skill references and installs dependencies from trusted sources, including the official AWS samples repository on GitHub and the @aws/agentcore package from the npm registry. These downloads are part of the intended migration and assessment functionality.
  • Command Execution: The discovery and generation phases involve executing shell commands such as gcloud for infrastructure inventory and curl for interacting with AI provider usage APIs. These operations are read-only, consent-gated, and necessary for the skill's primary purpose.
  • Data Exfiltration: Usage telemetry is collected during migration phases and sent to an AWS-owned endpoint (saws.activate.aws.dev). This process is documented as optional and is used to improve the tool. The transmitted data is anonymized and does not include sensitive information such as resource names or credentials.
  • Dynamic Context Injection: The skill uses shell command substitution in SKILL.md to check for the presence of local development tools like uv and uvx. This is a standard practice for environment validation and is implemented using benign version-check commands.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 19, 2026, 12:32 AM