wizard
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DYNAMIC_EXECUTION]: The skill's primary function is to generate executable bash scripts based on a provided template (template.sh). These scripts are designed for manual execution by a human to automate infrastructure provisioning, credential setup, and environment configuration. The library uses standard system tools for cross-platform URL opening and secret management.
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read local repository files, such as READMEs, .env files, and GitHub workflow configurations, to identify the variables and secrets required for the setup wizard. This represents a potential ingestion surface for untrusted data stored within the environment.
- Ingestion points: Local files like .env, .env.example, README, and .github/workflows/* are processed to scope the wizard's requirements.
- Boundary markers: No specific delimiters are defined in the instructions to isolate the content of analyzed files from the agent's logic.
- Capability inventory: The skill generates scripts that can write to the local file system (.env), interact with the GitHub CLI (gh), and initiate browser operations.
- Sanitization: The instructions focus on extracting key names and manual procedures; the generated script logic remains controlled by the static template.sh library.
Audit Metadata