Skip to content
For Managed AI Platform users, see the cloud platform documentation
Agentic AI > Build > DataRobot agentic skills

DataRobot agentic skills

DataRobot agentic skills are modular capability packages that help a coding agent execute DataRobot workflows reliably. Each skill bundles instructions, examples, and supporting resources in a SKILL.md file so the agent loads only what it needs for the current task.

Skills are maintained in the DataRobot Agentic Skills repository and work with Claude Code, Cursor, Codex, Gemini CLI, VS Code Copilot, OpenCode, and other coding agents.

Section Description
Quick start Install skills, verify the installation, and configure the environment.
Install from agent marketplaces Recommended install path for Claude Code, Cursor, Gemini CLI, and Codex.
Install with the universal installer Install all skills or target a specific skill or agent with one command.
Use skills in a coding agent Prompt examples and helper scripts.
Available skills Full skill catalog grouped by workflow.
Next steps Continue with Agent Assist or the Agentic Starter template.

Quick start

Complete these steps in order on first install:

  1. Install from a marketplace for Claude Code, Cursor, Gemini CLI, or Codex, or install with the universal installer for other agents.
  2. Verify installation by asking the coding agent which DataRobot skills are available.
  3. Configure the environment by running the datarobot-setup skill once per workspace.
  4. Use a skill in a prompt, or continue to Agent Assist or the Agentic Starter.

Skills nomenclature

"Skills" is an Anthropic term used in Claude AI and Claude Code, but the concept applies more broadly. OpenAI Codex uses AGENTS.md to define agent instructions, and Gemini uses gemini-extension.json for extensions. The DataRobot skills repository is compatible with all of them, and more.

List of supported agents

Supported agents for DataRobot skills include Claude Code, Cursor, Codex, Amp, VS Code Copilot (GitHub Copilot), Gemini CLI, Goose, Letta, Kilo Code, OpenCode, Windsurf, and Devin.

Install from agent marketplaces

Install DataRobot skills from the coding agent marketplace or extension catalog. This is the recommended onboarding path for Claude Code, Cursor, Gemini CLI, and Codex.

Agent Install surface
Claude Code claude.com/plugins/datarobot-agent-skills
Cursor cursor.com/marketplace/datarobot
Gemini CLI geminicli.com/extensions
Codex developers.openai.com/codex/plugins

Select the tab for the coding agent in use:

Install all DataRobot skills from the official Claude plugins marketplace.

From the terminal:

claude plugin install datarobot-agent-skills@claude-plugins-official

From within a Claude Code CLI session:

/plugin install datarobot-agent-skills@claude-plugins-official

Alternative: register the repository as a plugin marketplace from within a Claude Code CLI session:

/plugin marketplace add datarobot-oss/datarobot-agent-skills

To install a specific skill:

/plugin install datarobot-model-training@datarobot-skills

Open the DataRobot marketplace entry and click Add to Cursor, or run the following in Cursor chat:

/add-plugin datarobot-agent-skills

When the DataRobot Agentic Skills repository is open as the workspace, Cursor also reads AGENTS.md and makes skills available without additional configuration. To verify that the skills are loaded, open the AI chat panel (Cmd/Ctrl + L) and ask: "What DataRobot skills are available?"

This repository includes gemini-extension.json for Gemini CLI integration. Install from the GitHub URL:

gemini extensions install https://github.com/datarobot-oss/datarobot-agent-skills.git --consent

Or install locally after cloning the repository:

gemini extensions install . --consent

For more information, see the Gemini CLI extensions documentation.

Codex identifies skills through the AGENTS.md file in the workspace. Install skills with the universal installer, targeting Codex:

npx ai-agent-skills install datarobot-oss/datarobot-agent-skills --agent codex

Verify that the instructions are loaded:

codex --ask-for-approval never "Summarize the current instructions."

For plugin discovery and installation in ChatGPT and Codex, see the Codex plugins documentation. For details on how Codex reads agent instructions, see the Codex AGENTS.md guide.

Install with the universal installer

Install all DataRobot skills, or only the ones needed, for all supported AI agents with one command by using the universal skills installer.

To install all skills:

npx ai-agent-skills install datarobot-oss/datarobot-agent-skills

To install a specific skill:

npx ai-agent-skills install datarobot-oss/datarobot-agent-skills/skills/datarobot-predictions

To target a specific agent:

npx ai-agent-skills install datarobot-oss/datarobot-agent-skills --agent cursor
npx ai-agent-skills install datarobot-oss/datarobot-agent-skills --agent claude

Default behavior

By default, the installer copies skills to all supported agents at the same time. No configuration is required. For agent-specific installation methods, see Install from agent marketplaces.

Install on other coding agents

Use the following methods to install DataRobot skills on additional supported coding agents.

VS Code Copilot (GitHub Copilot)

VS Code with GitHub Copilot automatically detects and uses skills from this repository through the AGENTS.md file.

  1. Open this repository in VS Code.
  2. Ensure that the GitHub Copilot extension is installed and activated.
  3. Skills are automatically available through the AGENTS.md file.

Open Copilot Chat (Cmd/Ctrl + I) and ask "What DataRobot skills are available?" to verify that the skills are loaded.

Tip

The @workspace agent in Copilot Chat provides full context about the repository and available skills.

OpenCode

Add to ~/.config/opencode/opencode.json:

{
  "plugin": ["opencode-datarobot-skills"]
}

OpenCode automatically installs the plugin on startup. For more information, see the DataRobot OpenCode plugin.

Windsurf and Devin

Windsurf and Devin support the Agent Skills format. Install DataRobot skills using the universal installer.

Verify installation

After installing, ask the coding agent:

What DataRobot skills do I have available?

The response includes skills such as datarobot-setup and datarobot-agent-assist.

Configure the environment

Before using platform-connected skills, run the datarobot-setup skill once per workspace. It checks Python and other dependencies, authenticates dr-cli, and optionally adds DATAROBOT_ENDPOINT and DATAROBOT_API_TOKEN to the shell configuration.

Trigger setup with either:

  • Run datarobot-setup
  • /datarobot-setup

Use skills in a coding agent

After a skill is installed, mention it directly in instructions to the coding agent:

  • "Use the DataRobot model training skill to create a new project and start AutoML training."
  • "Use the DataRobot predictions skill to generate a prediction dataset template for deployment abc123."
  • "Use the DataRobot feature engineering skill to analyze feature importance for my model."
  • "Use the DataRobot model monitoring skill to check Data Drift for deployment xyz789."
  • "Use the DataRobot external agent monitoring skill to instrument my agent for DataRobot monitoring."

The coding agent loads the corresponding SKILL.md instructions and any helper scripts needed while completing the task.

Helper scripts

Some skills include helper scripts that an agent can run directly:

  • datarobot-predictions: get_deployment_features.py, generate_prediction_data_template.py, validate_prediction_data.py, make_prediction.py
  • datarobot-model-training: create_project.py, start_training.py, list_models.py
  • datarobot-model-explainability: compute_shap_matrix.py
  • datarobot-data-preparation: upload_dataset.py
  • datarobot-external-agent-monitoring: create_shell_deployment.py, verify_otel_connection.py
  • datarobot-agent-assist: select_framework.py, clone_template.py, setup_template.py, list_llm_models.py, rehearsal.py, env_utils.py

These scripts are located in each skill scripts/ directory and can be executed directly or used as references when writing code.

Available skills

The DataRobot Agentic Skills repository contains skills for common DataRobot workflows. Contributors can also add their own skills.

Start with these skills

  • datarobot-setup: Configure local development tools, authentication, and environment variables.
  • datarobot-agent-assist: Design, build, simulate, and deploy agents to DataRobot.
  • datarobot-discover: Find skills, MCP servers, agents, and platform resources.

Onboarding skills

The following skills support first-time setup and agent onboarding:

Skill Folder Description Documentation
skills/datarobot-setup/ Local DataRobot development setup (SDK, dr-cli, Agent Assist). SKILL.md
skills/datarobot-agent-assist/ Design, build, simulate, and deploy AI agents to DataRobot. SKILL.md
skills/datarobot-discover/ Discover DataRobot skills, MCP servers, agents, and platform resources. SKILL.md

Agent development

The following skills support agent design, deployment, and observability:

Skill Folder Description Documentation
skills/datarobot-agent-llm-selection/ Configure LLM integration for a DataRobot agent application and safely synchronize provider credentials. SKILL.md
skills/datarobot-app-framework-cicd/ Set up CI/CD pipelines for DataRobot application templates with GitLab and GitHub Actions. SKILL.md
skills/datarobot-external-agent-monitoring/ Instrument external agents with OpenTelemetry for DataRobot monitoring and observability. SKILL.md

ML workflows

The following skills support model training, deployment, predictions, and monitoring:

Skill Folder Description Documentation
skills/datarobot-model-training/ Instructions and utilities for training models, managing projects, and running AutoML experiments. SKILL.md
skills/datarobot-model-deployment/ Tools for deploying models, managing deployments, and configuring prediction environments. SKILL.md
skills/datarobot-predictions/ Guidance for making predictions, batch scoring, real-time predictions, and generating prediction datasets. SKILL.md
skills/datarobot-feature-engineering/ Instructions for feature engineering, Feature Discovery, and feature importance analysis. SKILL.md
skills/datarobot-model-monitoring/ Tools for monitoring model performance, tracking Data Drift, and managing model health. SKILL.md
skills/datarobot-model-explainability/ Tools for model explainability, Prediction Explanations, SHAP values, and model diagnostics. SKILL.md
skills/datarobot-data-preparation/ Utilities for data upload, dataset management, and data validation. SKILL.md

Platform and integration

The following skills support platform integration and workload management:

Skill Folder Description Documentation
skills/datarobot-workload-api/ Create, configure, debug, observe, and roll out container workloads on the Workload API. SKILL.md

Skill structure

Skills are self-contained folders that package instructions, scripts, and resources for a specific use case. Each folder includes a SKILL.md file with YAML frontmatter (name and description), followed by the guidance the coding agent uses while the skill is active.

Skill naming convention

All DataRobot skills follow the naming convention datarobot-<category>, where <category> describes the skill focus area. This provides clear identification of DataRobot-specific skills, consistent naming across the skill library, and easy discovery and organization.

Next steps

After installing skills and running setup, choose an onboarding path to create a first agentic workflow:

Topic Description
Start with Agent Assist Design, code, and deploy an agent with dr assist through natural conversation.
Get started with the Agentic Starter Install prerequisites, build, deploy, and test agentic workflows using DataRobot pre-built templates with dr start.
Agent Assist reference Workflows, environment variables, slash commands, and troubleshooting for Agent Assist.

See also