Get started

Arize AX for AI agents

Arize AX is an AI engineering platform for tracing, evaluating, and improving AI applications, used by ML teams to test prompts and models against datasets and monitor them in production. Connect it once through Arc0, and your agent, or Claude, ChatGPT and Cursor, can use it through one MCP endpoint, limited to what each user approved.

Connect Arize AXSee the 24 actions ↓
AI and modelsAPI keyMCP + RESTarize.com
AUDIT LOG · ARIZE AXPOLICY: acme-support
09:41:07 · claude · u_8f2read
arize_ax.list_spans
List Spans✓ allowed · 212ms
09:41:08 · claude · u_8f2read
arize_ax.get_space
Get Space✓ allowed · 164ms
09:41:09 · claude · u_8f2write
arize_ax.create_prompt
Create Prompt✓ approved · approved by user
EVERY ARIZE AX CALL, ON THE RECORD
01 · USE CASES

What agents do in Arize AX.

01

Run an experiment against a dataset

Create an experiment and add runs to evaluate a prompt against a dataset.

02

Review traces from a project

List traces and spans for a project to debug a model's behavior in production.

03

Publish a new prompt version

Create a new prompt version after testing changes through a completed experiment run.

02 · ACTIONS

24 Arize AX actions, graded by risk.

Every Arize AX action is tagged read, write or destructive, so one policy covers the whole app and new actions inherit the right default.

read

15

Look things up. Allowed by default.

  • arize_ax.get_space
    Get Space
  • arize_ax.get_prompt
    Get Prompt
  • arize_ax.list_spans
    List Spans
  • arize_ax.get_dataset
    Get Dataset
  • arize_ax.get_project
    Get Project
  • arize_ax.list_spaces
    List Spaces
  • arize_ax.list_traces
    List Traces
  • arize_ax.list_prompts
    List Prompts
  • arize_ax.list_datasets
    List Datasets
  • arize_ax.list_projects
    List Projects
  • arize_ax.get_experiment
    Get Experiment
  • arize_ax.list_experiments
    List Experiments
  • arize_ax.list_experiment_runs
    List Experiment Runs
  • arize_ax.list_prompt_versions
    List Prompt Versions
  • arize_ax.list_dataset_examples
    List Dataset Examples

write

9

Create and change things. Allow, or ask the user first.

  • arize_ax.create_prompt
    Create Prompt
  • arize_ax.update_prompt
    Update Prompt
  • arize_ax.create_dataset
    Create Dataset
  • arize_ax.update_dataset
    Update Dataset
  • arize_ax.create_experiment
    Create Experiment
  • arize_ax.add_experiment_runs
    Add Experiment Runs
  • arize_ax.add_dataset_examples
    Add Dataset Examples
  • arize_ax.create_prompt_version
    Create Prompt Version
  • arize_ax.update_dataset_examples
    Update Dataset Examples

destructive

0

Delete, cancel or archive. Ask first, or deny outright.

  • No destructive actions.
03 · HOW IT WORKS

Arize AX in three steps.

  1. 01Your users connect Arize AXThey add their Arize AX api key on Arc0 Connect, under your brand. It goes straight into the vault.
  2. 02You set the rulesReads run, and writes like “create Prompt” can wait for the user to approve.
  3. 03Any agent can actYour agent calls Arize AX through the Arc0 SDK or MCP, and so can Claude, ChatGPT and Cursor. Every call lands on the audit log.
POLICY.TS
await arc0.policies.set('arize_ax', {
  read: 'allow',
  write: 'ask',        // create_prompt
  destructive: 'deny',  
})

# Claude Code: the same connection, one URL
$ claude mcp add --transport http arc0 \
    https://mcp.arc0.ai/u/u_8f2
04 · AUTH AND DATA

How Arize AX connects.

Users add their Arize AX api key on Arc0 Connect. It is encrypted in the vault, never shown to the model, and each user can rotate or revoke it at any time.

The same Arize AX connection serves your agent over MCP and your own backend over REST and the proxy, so a user connects once. How Arc0 handles credentials →

AUTH
API key
CREDENTIALS
Per-tenant encrypted vault
MODEL SEES
Results only, never credentials
AUDIT LOG
Every call, on every plan
05 · WORKS WITH

Use Arize AX from any agent.

Claude
ChatGPT
Cursor
Codex
VS Code
OpenAI Agents SDK
Claude Agent SDK
Vercel AI SDK
Mastra
LangGraph
06 · RELATED

More ai and models apps.

Perplexity AI9 actionsAffinda119 actionsAgent Mail5 actionsAI/ML API30 actionsAivoov1 actionsAll Images AI7 actionsAltText.ai10 actionsAltTextLab2 actions
ALL AI AND MODELS APPS →
07 · FAQ

Arize AX and Arc0, answered.

Q01

Can I use Arize AX with Claude, ChatGPT or Cursor?

Yes. Connect Arize AX to Arc0 once, then add your Arc0 MCP URL to Claude, ChatGPT, Cursor, Claude Code or any other remote-MCP client. Each assistant only gets the Arize AX actions you allow.

Q02

How do users connect Arize AX?

Users add their Arize AX api key on Arc0 Connect. It is encrypted in the vault, never shown to the model, and each user can rotate or revoke it at any time.

Q03

Which Arize AX actions can my agent take?

24 in total: 15 read, 9 write and 0 destructive, such as “create Prompt”. Your policies decide which of them each agent may call.

Q04

Can I make my agent read-only in Arize AX?

Yes. Allow read actions and deny writes in the Arize AX policy. Your agent can still look things up, and any write it attempts is blocked and logged.

Q05

Can my own backend call Arize AX too?

Yes. The same Arize AX connection is available over REST and through the Arc0 proxy, so your product and your agent share one connection per user.

Get started

Plug Arize AX into your agent.

Your users connect Arize AX once, under your brand. Your agent gets 24 actions behind your policies, with every call on the record.

Connect Arize AXBrowse all apps →

Free to build · MCP + REST · Audit log on every plan