39 docs tagged with "fusion-ai"

Enabling Fusion AI for a tenant, the settings and policies that control it, using the Fusion chat in your own extensions, and publishing library playbooks

Fusion AI agents replay a set of questions on demand, on a schedule or when an automation fires

REST API for the Fusion AI agent service

These variables hold the Fusion AI configuration that you can create and save from ODSL: agents, custom assistants and custom tools. They are typed objects, so you set their properties like any other object. See Fusion AI for what each one does.

REST API for the Fusion AI assistant service

REST API for the Fusion AI playbook service

REST API for the Fusion AI playbook run service

REST API for the Fusion AI conversation service

REST API for the Fusion AI playbook task service

REST API for the Fusion AI custom tool service

REST API for the Fusion AI token usage service

The built-in Fusion AI assistants, and creating custom assistants with your own prompt, model and effort level

Full reference of the tools exposed by the Fusion AI MCP server

How Fusion AI tokens are billed, the Fusion AI Billing insight, and setting a monthly budget, limit and notifications

Use the Build an extension playbook to plan, write and save a complete extension from a description of its tabs, with an approval before anything is saved

Verifying each step with checks, approvals with gates, and choosing the next step with transitions and conditions

Add the Fusion AI MCP server to ChatGPT as a custom connector

Add the Fusion AI MCP server to Claude Desktop or Claude.ai

How to create a Fusion AI playbook, from a description, from the library, from an example or by hand

Fusion AI can create Smart Curves directly on the platform on your behalf — no ODSL scripting required. You describe the curve you need in plain language, Fusion AI confirms the details, and then creates it for you using its built-in Smart Curve tool.

Connect Fusion AI to your own systems, databases and APIs with custom tools written as ODSL or JavaScript functions

Describe a composition in your own words and let Fusion AI find the data and design it, or ask it to change one

A complete annotated playbook, and the flow of each playbook in the OpenDataDSL library

REST API for the Fusion AI feedback service

Ask Fusion AI to propose a data loader's event type, mapping, conditions, event id and dataset id from the sample

The Fusion AI insights for agent logs, feedback, playbook runs and token usage

What the Fusion AI MCP server is, what it lets you do, and how it's secured

The Fusion AI playbook format in detail, covering front matter, prose, inputs, steps, outputs and references

A one-page summary of the playbook format, and what the editor's error and warning messages mean

Filling in and running playbook tasks, background runs, schedules, approvals, resuming runs and monitoring

Example prompts to try with the Fusion AI MCP server

How the device sign-in flow works for the Fusion AI MCP server

Have Fusion AI design a smart curve or timeseries from a description, adapt a library example, change one, fix a failed preview or write a script function

Chat with Fusion AI in the Studio: conversations, choosing an assistant, tool activity, questions Fusion asks you, feedback, and saving a conversation as an agent

The OpenDataDSL playbook library, copying and customising playbooks, sharing playbooks in extensions, and running playbooks from a Fusion chat

Describe a downstream system to Fusion AI, answer the questions it asks, and apply its proposal to create a custom system or change an existing one

An introduction to Fusion AI Playbooks, repeatable AI workflows that run the same way every time

An introduction to Fusion AI, the OpenDataDSL AI assistant, and the Fusion AI extension in the portal

Ask Fusion AI to write, change, explain or fix a script from the editor, review its change side by side and apply it to the editor