AI agents

AI agents help you develop DataMiner solutions using task-specific instructions, reusable skills, and tools. You describe what you need, and the agents can inspect files, make changes, and run checks in your workspace.

Skyline distributes them through use-case-specific plugins in the AI Agent marketplace. They run in your AI client, such as GitHub Copilot or Claude Code. They should not be confused with "DataMiner Agents", i.e., DataMiner nodes.

Agents, skills, and plugins

• An agent has a specialized role, such as implementing, investigating, or reviewing changes.
• A skill provides guidance and resources for a specific task.
• A plugin bundles the agents and skills needed for a use case.

A plugin can contain one agent or several cooperating agents. An orchestrator can coordinate the workflow and delegate tasks to specialized agents.

Why use an agent

AI agents bring DataMiner-specific guidance into your workspace, reduce repetitive work, and help you follow established development patterns. A plugin brings the relevant expertise together so you do not need to configure each agent and skill separately.

When to use an agent

Choose a plugin that matches your use case, such as application development, connector development, or automation scripting. Check the AI Agent marketplace for currently available plugins and their scope. If a plugin provides an orchestrator, use it as the entry point to coordinate the work.