connect-recommend
Pass
Audited by Gen Agent Trust Hub on Jun 17, 2026
Risk Level: SAFE
Full Analysis
- Codebase and Input Analysis: The skill analyzes user-provided business information (URLs and descriptions) and local project files to generate recommendations. This process allows the agent to provide customized advice based on the existing code and stated requirements.
- Ingestion Points: Information enters the system through direct user input and automated file scans (SKILL.md Steps 1 and 2).
- Functional Capabilities: The skill is authorized to read project files, use search tools like
grep, and write a recommendation plan (connect-recommend-plan.md) to the project directory. - Task Delegation: It utilizes a sub-agent for business research, which is a standard architectural pattern for complex analysis tasks.
- Vendor Resource Integration: The skill references official configuration resources and directs users to official documentation for pricing and account management, ensuring that users receive information from verified sources.
Audit Metadata