datarobot-agent-skills Plugin
Plugin Claude Code Development AI Agents & AI App DevelopmentClaude Code Customization & WorkflowMonitoring & ObservabilityThe explanation below is AI-generated. Please verify it against the sources.
datarobot-agent-skills is a GitHub repository that packages modular "Agentic skills" (Agent Context Protocol definitions) for DataRobot AI/ML workflows, covering model training, deployment, predictions, feature engineering, monitoring, explainability, data preparation, App Framework CI/CD, and external agent monitoring. Per the README, it can be installed with the command npx ai-agent-skills install datarobot-oss/datarobot-agent-skills and is compatible with coding agents such as Claude Code, OpenAI Codex, Google Gemini CLI, Cursor, VS Code Copilot, Amp, Goose, Letta, Kilo Code, and OpenCode. Each skill is a self-contained folder with a SKILL.md file (YAML frontmatter plus guidance) and, in some cases, helper Python scripts that the agent can run directly. The repository also documents agent framework integration for LangGraph, PydanticAI, CrewAI, and LlamaIndex, and includes contribution guidelines. The homepage source is mostly marketing navigation content about DataRobot's broader Agent Workforce Platform and is not specific to this skills repository.
Overview
According to its homepage, DataRobot describes itself as a "Unified Agent Workforce Platform for Enterprise" that helps organizations build, operate, and govern AI agents, spanning agentic AI, generative AI, predictive AI, AI governance, and AI observability. The datarobot-agent-skills repository is a separate open-source project that provides skill packages for coding agents to interact with DataRobot's AI/ML workflows.
What you can do with datarobot-agent-skills
- Train models and run AutoML experiments via datarobot-model-training
- Deploy and manage models and prediction environments via datarobot-model-deployment
- Generate prediction datasets and run batch or real-time predictions via datarobot-predictions
- Perform feature discovery and feature importance analysis via datarobot-feature-engineering
- Monitor model performance and track data drift via datarobot-model-monitoring
- Compute SHAP values and prediction explanations via datarobot-model-explainability
- Upload and validate datasets via datarobot-data-preparation
- Set up CI/CD pipelines for DataRobot application templates (GitLab/GitHub Actions) via datarobot-app-framework-cicd
- Instrument external agents (Google ADK, LangChain, LangGraph, CrewAI, LlamaIndex, PydanticAI, generic Python) with OpenTelemetry via datarobot-external-agent-monitoring
- Build and deploy agents (LangGraph, CrewAI, LlamaIndex, NAT, Base) bundled with MCP server, backend APIs, and React frontend via datarobot-agent-assist
Sources
Original description (English)
DataRobot skills for AI/ML workflows — model training, deployment, predictions, feature engineering, monitoring, explainability, data preparation, App Framework CI/CD, and external agent monitoring.
History of datarobot-agent-skills
- Plugin Added datarobot-agent-skills