AI and media

AI agents that use OpenAI

UtilitySpaces gives an AI agent scoped access to OpenAI through its own encrypted connection, not your password. The agent can call 3 named OpenAI tools, and the one that writes waits for a person's approval unless an admin has set the connection to run without asking. Every call is written to a hash-chained audit log with its arguments and its result.

What an agent can do in OpenAI

List available models and generate chat completions with an OpenAI API key. An agent calls these by name, and matching is exact.

Every tool a OpenAI connection exposes, and whether it writes, which waits for approval by default
Tool	What it does	Waits for approval by default
List models	List models available to this key.	No — read only
Get model	Get one model's metadata.	No — read only
Create chat completion	Generate a chat completion (billed API call).	Yes — it writes

2 read tools, 1 write tool. Generated from the connector catalogue the product runs on, so it cannot list a tool that does not exist.

What teams hand an agent in OpenAI

Teams hand an agent an OpenAI key when a flow needs a model call from their account: checking what is available with List models, generating text with Create chat completion. That completion is marked as a write because it is a billed API call, so it waits for approval unless an admin has set this connection to run without asking.

Check which models the key can reach before a flow is built, by listing models and reading one model's metadata.
Generate a draft, such as a summary or a translation, with a specific model when a flow step calls for it (after a person approves it)
Run a one-off classification of a message through a chosen model when the inbox needs a second opinion on it (after a person approves it)

Before you let it write, or keep out: Do not put Create chat completion inside a loop the agent controls: every call bills the connected key, and on a connection set to run without asking nothing stops a batch from running.

What waits for a person

Reads run on their own. These do not — each one waits in Approvals for an owner or admin unless an admin has set this connection to run writes without asking, and every call toOpenAI, approved, rejected or unattended, is logged.

Create chat completion POST

What connecting OpenAI takes

OpenAI uses an API key you create in OpenAI and paste in once. A workspace admin sets the connection up; the fields are:

API Key

Credentials are encrypted and bound to your workspace, never read back, and every outbound call passes one guarded fetch — the connector security posture sets out exactly how, and the security page covers the platform around it.

Where OpenAI fits

Give an agent OpenAI work by writing a message in the inbox, or put its tools in a scheduled flow. What the run cost is on the cost breakdown.

Other ai and media integrations: AnythingLLM, AssemblyAI, D-ID. Or see every application agents can connect to.

See it doing your actual work

Open a demo workspace, connect one app, and give an agent a real job. No card, no sales call.

Try the demo See pricing

We reply to every message within one business day.