Feature Request: Security Checker for Vibe Coders & Agent Import/Orchestration in AI Toolkit

1. Security Checker for Vibe Coders

Problem:
Many "vibe coders" (beginner or casual developers) are often unaware of best security practices, which leads to frequent issues such as unintentionally exposing API keys or misconfiguring sensitive files. There have been numerous cases where public repositories contained secrets or insecure configurations.

Proposed Solution:
Develop a security checker tool that allows users to upload their project repositories. The tool should automatically analyze the repository for common security misconfigurations, including:

• Exposed API keys or secrets in code, configuration files, or environment files.
• Insecure default settings (e.g., open CORS or weak passwords).
• A missing .gitignore file for sensitive files.
• Publicly accessible endpoints without authentication.
• Outdated or vulnerable dependencies.

This tool would serve as a secondary check for developers, particularly those who are less familiar with security, helping to prevent accidental leaks or vulnerabilities.

2. Script-to-Agent Import & Orchestration in AI Toolkit

Problem:
Currently, when using the AI Toolkit (AITK), only a limited number of agents (such as the GitHub MCP server) can be added. Other agents (e.g., HackathonAgent, EventAgent, GithubAgent) created using Semantic Kernel (SK) scripts are not accessible through the agent builder or playground. This limitation restricts the ability to orchestrate and test multiple agents together.

Proposed Solution:
Implement a feature that reads SK scripts and automatically creates corresponding agents in the AITK based on the script's instructions and names. These agents should then be accessible in the agent builder and playground, allowing users to orchestrate interactions between them and test workflows across different models. Additionally, enable the importing and management of custom agents created with SK, so they can be used alongside built-in agents like the GitHub MCP server.

Example Use Case:
A user should be able to add multiple agents (e.g., HackathonAgent, EventAgent, GithubAgent) to the AITK, orchestrate their interactions, and test these workflows in the playground, not just with the GitHub MCP server.se:**
A user should be able to add multiple agents (e.g., HackathonAgent, EventAgent, GithubAgent) to the AITK, orchestrate their interactions, and test these workflows in the playground, not just with the GitHub MCP server.

Reference:
See my GitHub commit for context on current limitations and user workflow.