DataRobot Agent Skills and MCPs are now discoverable through Agentic Resource Discovery

2026-06-17 · Source: Blog | DataRobot · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems · Depth: Intermediate, short

Summary

DataRobot has announced support for the Agentic Resource Discovery Specification (ARD), an open standard designed to simplify the discovery and verification of agentic resources across the web. This integration makes DataRobot Agent Skills and MCPs (Model Control Planes) discoverable by AI clients, registries, and developers through a standard ".well-known/ai-catalog.json" path, specifically at https://datarobot.com/.well-known/ai-catalog.json. ARD addresses the previous manual process of wiring agents to capabilities, enabling a shift from static integrations to dynamic discovery. DataRobot's ARD catalog currently includes skills for model training, deployment, predictions, feature engineering, monitoring, explainability, data preparation, App Framework CI/CD, external agent monitoring, and Agent Assist, providing agents with operational context for DataRobot workflows. This initiative aims to make enterprise AI agents easier to build, operate, monitor, and govern by ensuring platform context is discoverable.

Key takeaway

For AI Engineers building agents on DataRobot, you no longer need to manually configure every skill and MCP. DataRobot's ARD support means your agents can dynamically discover necessary platform context, streamlining development and integration. If you are governing enterprise agentic AI, you gain a standard way to control which catalogs and registries your agents access, enhancing governance. Explore the ARD specification and DataRobot's quickstart guide to utilize this new capability.

Key insights

DataRobot now uses ARD to make its Agent Skills and MCPs dynamically discoverable, moving beyond manual agent integrations.

Principles

• Agent utility depends on discoverable capabilities.
• Dynamic discovery replaces static agent wiring.
• Agents require operational context, not just docs.

Method

Providers publish a catalog of resources (skills, MCPs, APIs) under their domain. Discovery services and AI clients then find, index, and resolve these resources when an agent needs them.

In practice

• Publish an AI catalog via ARD specification.
• Create DataRobot Agents with discoverable skills.

Topics

• Agentic AI
• Agentic Resource Discovery
• DataRobot Agent Skills
• Model Control Plane
• AI Catalogs
• Enterprise AI Governance

Code references

• ards-project/docs

Best for: CTO, VP of Engineering/Data, Director of AI/ML, AI Engineer, MLOps Engineer, AI Architect

Related on AIssential

• AI agents are getting their own search engine — Agentic Resource Discovery (ARD) creates a standardized "search engine" for AI agents to find and verify capabilities.
• DataRobot for Developers: Skills, MCP, and the agentic developer surface — DataRobot simplifies AI agent development by integrating core services directly into developer workflows.
• Agent Skills — Part 2: Bridging Skills with Production Tool Ecosystems — SkillRegistry decouples tool discovery from execution, enabling scalable agent skill management with existing production tools.
• DataRobot Q4 update: driving success across the full agentic AI lifecycle — Transitioning AI agents from prototype to production requires robust orchestration, governance, and scalable infrastructure.
• Bridging Design Systems and Code with MCP and AI Agents — Design systems, context engineering, and agentic AI combine to enable AIs to build consistent software by following explicit rules via MCP.
• Skills, Tools and MCPs - What’s The Difference? — AI agent capabilities evolve from basic function calls to standardized protocols and explicit expertise layers.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Blog | DataRobot.