create-agent-template
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied agent descriptions and interpolates them into generated template files without specific sanitization instructions.\n
- Ingestion points: User-provided descriptions of the agent's intended purpose and tasks (SKILL.md).\n
- Boundary markers: The generation logic lacks explicit boundary markers or instructions to the model to ignore potentially malicious instructions embedded in user data.\n
- Capability inventory: The generated pipeline templates include shell script execution (
bash) and container-based plugin steps that interact with repositories and external APIs (references/agent-examples.md).\n - Sanitization: No specific instructions are provided for the AI to sanitize or escape user input before interpolation into JSON or YAML structures.\n- [DYNAMIC_EXECUTION]: The skill generates shell scripts and YAML configuration files that are executed within a CI/CD environment.\n
- Script generation: Produces bash scripts for repository interaction, SCM provider detection, and Git operations, utilizing platform-specific environment variables for secure credential handling.
Audit Metadata