Agent Native

databricks-ai-runtime

This skill enables agents to submit and manage GPU training workloads on Databricks serverless compute using the `databricks air` CLI.

by DatabricksOfficial publisherNot reviewed

Identity proven by GitHub's verified organisation badge (databricks)

Agents
Claude Code, Codex, Cursor, Gemini CLI, GitHub Copilot, OpenCode, claude.ai and API
License
LicenseRef-Databricks (restricted)
Pin
e3df77bpinned Oct 9, 2026
Context cost
about 2,900 tokenswhen the skill loads
GitHub stars
345on databricks/databricks-agent-skills
Scripts
None found

What it does

“Databricks AI Runtime, the `databricks air` CLI commands for submitting and managing GPU training workloads on Databricks serverless compute. Use for: writing and submitting `databricks air` workload YAML, passing hyperparameters and secrets, checking run status, listing/cancelling runs, streaming a run's logs and watching its progress, custom Docker image setup, and environment configuration.”
The author's description, from SKILL.md

Generated summary from the skill's own text

This skill enables agents to submit and manage GPU training workloads on Databricks serverless compute using the `databricks air` CLI. It covers writing workload YAML configurations, passing hyperparameters and secrets, monitoring run status, streaming logs, and handling distributed training across multiple nodes with various frameworks like PyTorch and Hugging Face Accelerate.

When to use it

  • Submit distributed GPU training jobs without managing infrastructure
  • Configure multi-node training with specific accelerator types and environment variables
  • Pass hyperparameters and secrets to training scripts via YAML
  • Monitor active training runs and stream their logs in real time

When not to

  • Databricks CLI with `air` command is not installed or unavailable
  • Managing small-scale single-GPU experiments on managed clusters
  • Submitting non-training batch workloads or non-GPU compute tasks

Install

Restricted license: install from the publisher's repository and read its terms.

These commands install from the publisher's repository and need no account.

skills CLI

npx skills add databricks/databricks-agent-skills --skill databricks-ai-runtime

Installs from the repository's default branch, which may have changed since the commit we checked.

GitHub CLI 2.90 or later

gh skill install databricks/databricks-agent-skills databricks-ai-runtime@e3df77b3f0eb5d70392eb35c6761ca0f2929a4f6

Installs the commit we checked (e3df77b).

Copy by hand

Download the folder at e3df77b and put it in .claude/skills/databricks-ai-runtime for Claude Code and in .agents/skills/databricks-ai-runtime for Codex, Cursor, Gemini CLI, GitHub Copilot and OpenCode.

Skill folder on GitHub at e3df77b

Claude Code plugin

/plugin install databricks@claude-plugins-official

The publisher's plugin in Claude's official marketplace. It may hold more than this skill, and we have not pinned it.

Copy for your agent

A prompt for any coding agent. It installs this skill from GitHub at the pinned commit, copies the files as they are, and runs nothing unless you ask.

Prompt

Install the agent skill "databricks-ai-runtime" by Databricks.

Source, pinned to commit e3df77b3f0eb5d70392eb35c6761ca0f2929a4f6:
https://github.com/databricks/databricks-agent-skills/tree/e3df77b3f0eb5d70392eb35c6761ca0f2929a4f6/experimental/databricks-ai-runtime

1. Download that folder at exactly this commit. Do not use the default branch.
2. Copy every file unchanged, keeping the folder structure, into .claude/skills/databricks-ai-runtime/ for Claude Code and .agents/skills/databricks-ai-runtime/ for Codex, Cursor, Gemini CLI, GitHub Copilot and OpenCode. Create the folders if they do not exist.
3. Do not edit the files, and do not run any script in the folder unless I ask you to.
4. When you are done, list the files you copied and where you put them.

Its license is restricted or not stated. Show me the license terms from the repository before you copy anything.

Listing and checks: https://www.agentnative.dev/skills/databricks/databricks-ai-runtime

SKILL.md at the pin

The license of this skill does not let us show its files here.

Files

5 files, 43 KB in total. Each file has a sha256 in the pin's manifest; the first 12 characters are shown.

Skill folder on GitHub at e3df77b
All 37 skills from Databricks

More from Databricks

  • databricks-agent-bricks

    This skill enables agents to create and manage Knowledge Assistants for document question-answering via RAG on Volumes, and Supervisor Agents that orchestrate…

    Claude Code, Codex, Cursor +3 · about 1.7K tokens · pinned 6 days ago

  • databricks-ai-functions

    This skill provides built-in Databricks AI Functions that add AI capabilities directly to SQL and PySpark pipelines, including classification, extraction…

    Claude Code, Codex, Cursor +3 · about 3.8K tokens · pinned 6 days ago

  • databricks-aibi-dashboards

    Creates Databricks AI/BI dashboards using JSON specifications with support for multiple visualization types.

    Claude Code, Codex, Cursor +3 · about 9.5K tokens · pinned 6 days ago

  • databricks-app-design

    This skill helps design the user experience of custom-code Databricks Apps built with AppKit and React that display data.

    Claude Code, Codex, Cursor +3 · about 1.6K tokens · pinned 6 days ago

  • databricks-apps

    Build custom interactive data apps on Databricks using AppKit (TypeScript/React framework) or other frameworks like Streamlit or FastAPI.

    Claude Code, Codex, Cursor +3 · about 5.5K tokens · pinned 6 days ago

  • databricks-apps-python

    Enables agents to build Python backends for Databricks Apps using FastAPI, Flask, Dash, Streamlit, Gradio, or Reflex.

    Contains scripts

    Claude Code, Codex, Cursor +3 · about 2.9K tokens · pinned 6 days ago

All Cloud and DevOps skills
  • debug-openshell-cluster

    Diagnoses OpenShell gateway deployment problems across Docker, Podman, Kubernetes, and VM compute platforms.

    by NVIDIAOfficial publisher

    Claude Code, Codex, Cursor +4 · about 19.1K tokens · pinned 6 days ago

  • huggingface-spaces

    Create and deploy machine-learning applications on Hugging Face Spaces using Gradio, Docker, or Static SDKs.

    by Hugging FaceOfficial publisher

    Claude Code, Codex, Cursor +4 · about 4.3K tokens · pinned 6 days ago

  • upgrading-mwaa-environments

    Upgrades an MWAA environment to a newer Apache Airflow version by computing the optimal version-jump path, selecting a deployment approach (new environment…

    by Amazon (AWS)Official publisher

    Claude Code, Codex, Cursor +4 · about 7K tokens · pinned 6 days ago

  • deploying-airflow

    This skill guides deploying Airflow DAGs and projects using Astro (Astronomer's managed platform), open-source Airflow on Docker Compose, or Kubernetes with…

    by AstronomerOfficial publisher

    Claude Code, Codex, Cursor +4 · about 2.7K tokens · pinned 6 days ago

  • deploying-go-sdk-bundles

    This skill builds, packs, and deploys Go SDK bundles as self-contained AFBNDL01 executable artifacts for Airflow's ExecutableCoordinator to run.

    by AstronomerOfficial publisher

    Claude Code, Codex, Cursor +4 · about 1.7K tokens · pinned 6 days ago

  • deploying-java-sdk-bundles

    This skill teaches how to build Java SDK bundles as JAR files using Gradle or Maven, then deploy them to Airflow workers so the coordinator can execute Java…

    by AstronomerOfficial publisher

    Claude Code, Codex, Cursor +4 · about 2.6K tokens · pinned 6 days ago

Open roles at Databricks

The skills listed here are third-party software, collected from public GitHub repositories as a catalog of recent skills and provided as is. Agent Native does not write or control them and is not responsible for any harm, loss or damage caused by installing or using them. Labels and checks describe what we saw at one commit and are not a guarantee. Read a skill's files before you run it. Terms of use