Agency
Files
Documentation
Connectors
Contact Us
Hurozo CLI
Hurozo SDK

Worker Tools

Tools

Worker View

No tools selected

Team worker Untitled Agent Describe what this agent should do. HurozoAgent

Worker Chat

Describe the worker you want to add to this team, or what you want to change.

Draft JSON

{}

Teams

Teams are managed from Agency.

Connectors

Connect services for the current agency

Loading connectors…

Agency

Team workspace

Planning

Build

Runs

Select a run to inspect its activity.

Files

Documentation

Contact Us

Hurozo CLI

Terminal-first agent work

Hurozo CLI

The same Hurozo agent runtime, model routes, approvals, and connected tools from your shell.

Install with Homebrew

brew tap hurozo/hurozo-cli
brew install hurozo-cli

View latest GitHub release

Model routing

Switch between Hurozo, Gemini, Opus, and ChatGPT-backed routes when your account is configured for them.

Approvals

Review tool permissions from the terminal and keep remembered decisions scoped to the exact operation.

Second opinions

Use model arbitration to ask another available route for critique, checks, or a different perspective.

Connected tools

Use the same account-backed connectors and skills available in Hurozo's desktop and WebCLI experiences.

First run

After installation, run hurozo. The CLI walks you through account authentication and any model-specific sign-in or secret setup that your selected route requires.

Hurozo SDK

Python agent invocation

Hurozo SDK

Invoke saved Hurozo agents from Python using the same API tokens, inputs, outputs, and runtime routing used by Agent Builder.

Install

pip install hurozo

getting_started.pyPython

import json
import os

from hurozo import Agent

os.environ["HUROZO_API_TOKEN"] = "YOUR_API_TOKEN"
agent = Agent("sentiment-analyser")
inputs = {
    "prompt": "Analyze sentiment and return JSON. Text:",
    "input": "I am very happy!",
}

result = agent.run(inputs)
print(json.dumps(result, indent=2))

Agent lookup

Use an agent name, saved agent id, or agent UUID. The SDK resolves the correct execution target.

Structured inputs

Pass a Python dictionary matching the API input keys exposed by the saved workflow.

Hosted runtime

Runs against Hurozo's execution endpoint, including model routing, tools, outputs, and usage accounting.

Token setup

Create an API token from Account settings -> Manage API Tokens, then set it as HUROZO_API_TOKEN before running scripts.

Trust, security, and control

You reached security review. Here is the honest page.

This is the page you send to IT, risk, and legal. It splits what Attend proves today from what we are still building, in plain words, so nothing here surprises anyone in week six.

Your model, your keys EU AI Act, by design Every run leaves a trace

Two labels run through this page. Today means it works now and you can see it. Roadmap means it is planned or built on validated demand, not a promise you can act on yet. We do not dress a roadmap item as a fact.

The eight questions, and where each one stands

the short answers, in one table

The questionThe lineStatus

Residency As of August 2026 the application runs on Cloud Run in europe-west1 and primary data sits in Firestore in the EU (eur3). Model inference runs at the provider, global today. Today what is true today ↗

Who sees what Your team, by the role you set; the desk reads only the systems you connect and writes to none of them on its own. Today what is true today ↗

Logs and export Every action logged, replayable, and inspectable in the product. The packaged one-click export is on the roadmap, and says so. Today on the roadmap ↗

Human oversight Anything that spends money or breaks a policy is drafted and held at a gate you set; who said yes is on the record. Today what is true today ↗

The AI Act stance Designed around the record-keeping, transparency, and human-oversight concepts in Articles 12 to 14. Design intent, not a compliance assessment or certification. Today the documents ↗

DPA and GDPR A DPA on request, signed by a person; processing under GDPR with the residency above. Roadmap the documents ↗

The model provider Bring your own model account, or use ours. Inference runs at the provider's endpoint, and this page names the leg it does not cover. Today what is true today ↗

Retention Retention set per account; deletion on request, with the record exported first. Roadmap ask us ↗

Nothing here is a roadmap claim wearing a present tense. Every row is written out in full further down this page, and where we have not built a thing yet the row says Roadmap rather than rounding it up.

You have seen how this goes

the review that stalls

The region answer changed depending on who you asked.

Sales said EU. The doc said something else. Nobody could point to the setting.

The approval only existed in the demo.

"A human checks it" turned out to mean one screen in the pitch, not a real gate you control.

The audit log could not reconstruct the send.

When you asked to see exactly what went out and why, the trail had holes.

The promised document was a dead link.

The security overview, the DPA, the pen-test summary. Requested, then silence.

If you have run a vendor through review, you know these. So here is Attend split straight down the middle: what we prove today, and where we tell you plainly the work is not done.

What Attend proves today

marked Today, you can see it

Received

Refund, damaged on arrival