adk-sample-creator

Pass

Audited by Gen Agent Trust Hub on Oct 16, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [Standard Development Tool Usage]: The skill's instructions include a section for verifying code quality and correctness using established Go tools, such as go build, go vet, and go mod tidy, alongside the golangci-lint runner. These are common utilities in the Go ecosystem and are used here for project maintenance and validation.
  • [Configuration Management]: The guidelines promote reading configuration parameters, such as the GOOGLE_CLOUD_PROJECT, from environment variables rather than hardcoding them. This follows standard security practices for managing environment-specific settings and avoiding the exposure of configuration data in source code.
  • [Agent Input Surface]: The provided templates facilitate the creation of agents that interact with user prompts. As with any interactive system, the implementation of the agent's logic represents a surface for indirect prompt injection. Developers using these templates are encouraged to implement appropriate sanitization and safety guardrails within the agents they author.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 16, 2026, 11:05 AM