agentic-workflow-guide
Pass
Audited by Gen Agent Trust Hub on May 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill serves as a technical resource for building and managing AI agent workflows. It emphasizes security best practices, such as the principle of least privilege and the prevention of credential leakage, notably through the provided 'security.instructions.md' template.
- [COMMAND_EXECUTION]: The skill includes a local utility script, 'scripts/scaffold_workflow.py', which is used to automate the creation of directory structures for agent projects. Analysis of the script shows it performs standard file system operations (creating directories and writing template files) without any network access or dynamic code execution vulnerabilities.
- [EXTERNAL_DOWNLOADS]: The 'python.instructions.md' template mentions downloading 'uv', a well-known Python package manager, from 'astral.sh'. This is documented as a recommended development tool and does not constitute a security risk.
Audit Metadata