> ## Documentation Index
> Fetch the complete documentation index at: https://arize-ax.mintlify.site/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Skills

> Arize Skills teach your coding agent the full Arize AX workflow — tracing, datasets, experiments, evaluators, annotations, and prompt optimization — through natural language.

[Arize Skills](https://github.com/Arize-ai/arize-skills) encode the workflows we've refined building the Arize AX platform and helping teams debug LLM apps in production. They handle the `ax` CLI flags, data shape quirks, and multi-step recipes so your agent doesn't have to guess. Your agent gets the full Arize workflow — add tracing, export traces, create datasets, run experiments, set up evaluators, manage annotations, optimize prompts — all through natural language.

Works with Cursor, Claude Code, Codex, Windsurf, and [40+ other agents](https://github.com/vercel-labs/skills#supported-agents).

<Tip>
  **Already have traces?** Give your agent this prompt to install skills and start debugging:

  ```
  Install the Arize skills plugin from https://github.com/Arize-ai/arize-skills, then use the arize-trace skill to export and analyze recent traces from my project. Summarize any errors or latency issues you find.
  ```
</Tip>

## Next steps

<CardGroup cols={2}>
  <Card title="Install & authenticate" icon="download" href="/docs/ax/skills/install">
    Install the AX CLI, set up credentials, and add skills to your agent.
  </Card>

  <Card title="Skill catalog" icon="list" href="/docs/ax/skills/catalog">
    Every available skill and what it does.
  </Card>

  <Card title="CLI reference" icon="terminal" href="/docs/api-clients/cli/skills">
    Flags for `ax skills install` and `ax skills clear`.
  </Card>

  <Card title="Coding agent setup" icon="robot" href="/docs/ax/set-up-with-ai-assistants">
    The other agent paths — `npx evals`, Paste-and-Go, and MCP servers.
  </Card>
</CardGroup>

## Skills, MCP servers, or `npx evals`?

* **Skills** — you want your agent to run Arize workflows end to end: export traces, build datasets, run experiments, create evaluators, optimize prompts.
* **[MCP servers](/docs/ax/set-up-with-ai-assistants#mcp-servers)** — you want persistent instrumentation guidance and doc search in your IDE.
* **[`npx evals`](/docs/ax/set-up-with-ai-assistants#one-command-setup)** — you have no account or traces yet and want one command to do everything.

<Note>
  Looking for the **Skills tab** under **More → Agent Integrations**? Those are [Agent Improvement Loop skills](/docs/ax/agents/skills-and-permissions) — account-level integrations that give managed sandbox workers access to GitHub, Datadog, and Arize. Same idea, different runtime.
</Note>