modify-agent

Pass

Audited by Gen Agent Trust Hub on Feb 25, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references official Databricks repositories, including the databricks-langchain SDK and databricks-ai-bridge on GitHub. These are trusted vendor resources used for agent development.
  • [EXTERNAL_DOWNLOADS]: Includes an external reference to github.com/bbqiu/agent-on-app-prototype for prototype examples. This is presented as an informative resource for developers.
  • [COMMAND_EXECUTION]: Mentions the use of uv sync to manage the Python virtual environment and ensure the databricks_langchain package is correctly installed. This is a standard environment management command.
  • [PROMPT_INJECTION]: Provides architectural guidance on customizing agent behavior by prepending a system message (AGENT_INSTRUCTIONS) to the conversation history. This implementation follows security best practices by maintaining the separation between developer instructions and user-provided messages.
  • [DATA_EXFILTRATION]: Demonstrates how to connect to external services via MCPServer using generic placeholders for URLs and API keys. The examples are intended for tool orchestration and do not involve unauthorized data movement.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 25, 2026, 12:14 AM