How Builder fits with Conductor

Builder is for non-technical users who describe what they need in plain English. Conductor is for developers and power users who build flows visually on the canvas. Builder's output is a Conductor canvas — pre-populated by AI so your team can review or extend it without starting from scratch.

Step 1 of 7

What does this agent need to do?

Describe it in plain English. Don't worry about the technical details — that's our job.

Or pick a starting point

📸Daily briefing 📈CRM updater 📊Report generator 🎉Support router 📄Contract analyst 👥Hiring tracker

Step 2 of 7

What information will this agent need?

Where does your agent get its data? Tick all that apply.

📸Emails (Gmail) 👋Slack messages 📋Spreadsheet data 📄Documents / PDFs 📅Calendar events 📚Database records 🌐Webhook / API data 👆I tell it manually

Anything else?

Step 3 of 7

What should this agent produce when it's done?

Where should the results go? Tick all that apply.

👋Post to Slack 📸Send an email 📋Write to a spreadsheet 📚Update a database 📄Save to Notion 📈Update CRM / Salesforce 🌐Call a webhook / API 👤Show me a dashboard

Anything else?

Step 4 of 7

Which systems does this agent connect to?

We handle the connections — you just tell us which ones. You authorise access after launch.

📸

Gmail

👋

Slack

📋

Google Sheets

🛍

Shopify

📧

Klaviyo

🛡

WhoisXML

📷

Instagram

☁

Salesforce

📈

HubSpot

📚

Notion

📅

Calendar

🗺

PostgreSQL

🌐

Webhook

➕

Custom API

Step 5 of 7

What is this agent allowed to do without asking you?

You can change this anytime from your manage dashboard.

Full auto — just do it

Agent acts on everything within its role. Best for routine, low-risk tasks.

Auto with daily summary ✓

Runs automatically but sends a daily digest of everything it did. Recommended for most users.

Ask me for high-stakes actions

Handles routine work automatically, pauses before sending client emails, updating key records, or anything significant.

Always ask me first

Drafts a recommendation for every task and waits for your approval. Best for sensitive workflows.

Step 6 of 7

When should this agent run?

Pick the pattern that fits your workflow.

On a schedule ✓

Runs at specific times — e.g. every morning at 8am, or every Monday at 9am.

On demand — I'll trigger it

Runs when you click a button or make an API call. Best for one-off or batch tasks.

When triggered by an event

Runs instantly when something happens in another system — new lead, form submission, or incoming webhook.

Continuously — always on

Processes each incoming item in real time. Best for monitoring, live feeds, or always-on support agents.

Step 7 of 7

Give your agent a name, role and goal

The name and role appear on your workforce dashboard. The goal sets up performance tracking.

Agent name

Role title (2–4 words)

What does success look like? (one sentence)

Identity — how this agent appears across the platform

Named colleague

Has a person's name. Shows a portrait.

System agent

Called by API. Shows an icon, never a face.

Hire fee (USD) — one-time on hire

Per-task fee (USD) — per task completed

What happens next

Our AI reads your answers, selects the right agents from the marketplace, writes their system prompts, and assembles a complete Conductor canvas flow — ready for you to review, fund, and launch. No code required.

⚡

Assembling your agent...

Analysing your requirements and identifying agent roles.

Agent configured

Your agent is ready to review

Here's how we've set it up. Launch when you're happy — or open in Conductor to fine-tune the canvas.

Configuration

TriggerDaily at 8:00 AM

PermissionsAuto + daily summary

Probation period30 days

Trust tier requiredVerified

Estimated cost

Per run~85 EXC

Per month~2,550 EXC

Agents in flow4

IntegrationsGmail, Slack

Success goal — how we'll measure performance

Every morning before 9am you have a clear prioritised list of actions from your inbox and all deal updates are logged without you touching a thing.

Open in Conductor

List in Find
Public catalogue
EightX review: 24–48 hrs

Build to Forge
Your private workforce
EightX review: 24–48 hrs

Open in Conductor
Fine-tune canvas first
No review yet

Every agent is scanned and reviewed by EightX before it goes live — whether it's listed publicly on Find or built into your private Forge. The review covers security (Semgrep + Socket.dev), data handling, and runtime safety.

✓

Submitted for review

Your agent is submitted

EightX will review your agent within 24–48 hours. You'll receive an email when approved or if changes are requested. After approval, your agent enters 30-day probation mode where you can watch it run before promoting to live.

View my agents → Open in Conductor Build another