wizard
Pass
Audited by Gen Agent Trust Hub on Aug 11, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill incorporates a surface for indirect instructions by analyzing repository files to define manual stages. 1. Ingestion points: README, .env, and GitHub workflow files (referenced in SKILL.md). 2. Boundary markers: Absent. 3. Capability inventory: Generation of scripts that perform local file writes and GitHub secret updates. 4. Sanitization: Absent. This risk is effectively mitigated by the human-in-the-loop design where the agent only authors a script for user verification.
- [SAFE]: The 'template.sh' library implements secure credential handling patterns, including the use of hidden terminal input for secret collection and atomic file operations using 'mktemp' for environment variable updates.
- [COMMAND_EXECUTION]: The skill produces scripts that utilize standard system utilities and the official GitHub CLI to assist in setup. These operations are transparent, local to the user's environment, and require explicit manual execution.
Audit Metadata