> ## Documentation Index
> Fetch the complete documentation index at: https://docs.paradime.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Run DinoAI Background Agent

> Add a Run DinoAI Background Agent command to a Paradime Bolt schedule to give a DinoAI agent a task to run as part of your dbt™ pipeline.

Give a Paradime DinoAI agent a task to run as part of a Bolt schedule, so an autonomous agent runs on a cron schedule, on PR merge, on an upstream job's completion, or any other Bolt trigger.

Bolt schedules agents natively: when the trigger fires, Bolt dispatches the agent run directly inside your workspace. There is no API key, SDK, or CLI setup involved.

## Prerequisites

* A programmable agent, either defined under `.dinoai/agents/<name>.yml` in your repository or built in the app from the Agents page (see [Creating Agents in the App](/products/dino-ai/programmable-agents/creating-agents-in-the-app)).

<Info>
  **No API keys required.** The command runs natively inside your workspace, so there are no credentials to set on the schedule. Existing schedules that still set `PARADIME_API_ENDPOINT`, `PARADIME_API_SECRET`, or `PARADIME_WORKSPACE_UID` as environment variables keep working unchanged, and you can remove those variables.
</Info>

## Usage

Add the **Run DinoAI Background Agent** command to your schedule, then:

1. Select the **Agent name** from the dropdown. The list includes agents from the YAML files in `.dinoai/agents/` on the configured branch as well as agents built in the app.
2. Describe the **Task** you want the agent to complete (for example, "investigate why the run failed and propose a fix"). Be specific and reference concrete inputs like table names or time windows.
3. Click **Add command**.

<Tip>
  You can also start from the agent: on the **Agents** page of the DinoAI Agent app, hover an agent card and click **Schedule**. This opens a new Bolt schedule with the command already added and the agent pre-selected.
</Tip>

## What happens on a run

* Bolt dispatches the agent run natively and **waits for it**: the command blocks until the agent reaches a terminal state, and the Bolt run inherits the agent's success or failure.
* The agent's messages **stream into the command log**, so you can follow progress live in the Bolt run logs.
* The session also appears in the DinoAI Agent **sessions panel under the Bolt source**, so scheduled runs are easy to tell apart from API or Slack sessions.

<Info>
  Each run starts a fresh agent session with the configured task as its first message. It does not resume previous sessions.
</Info>

## Learn more

* [Programmable Agents overview](/products/dino-ai/programmable-agents/index)
* [Creating Agents in the App](/products/dino-ai/programmable-agents/creating-agents-in-the-app) for building an agent without writing YAML
* [Triggering from Bolt](/products/dino-ai/programmable-agents/triggering-from-bolt) for the full lifecycle, including schedules-as-code
* [YAML configuration](/products/dino-ai/programmable-agents/yaml-configuration) for authoring an agent under `.dinoai/agents/`
* [Tools reference](/products/dino-ai/programmable-agents/tools-reference) for the tools an agent can use


## Related topics

- [Creating Agents in the App](/products/dino-ai/programmable-agents/creating-agents-in-the-app.md)
- [Bolt runs agents natively, no API keys](/changelog/2026-08-05/bolt-runs-agents-natively.md)
- [Commands](/products/bolt/creating-schedules/command-settings/index.md)
- [Getting started with the DinoAI background agent](/guides/paradime-101/getting-started-with-the-dinoai-background-agent/index.md)
- [Live agent runs in the DinoAI app](/changelog/2026-07-30/live-agent-runs.md)