> For the complete documentation index, see [llms.txt](https://docs.lleverage.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.lleverage.ai/run-operations/agent.md).

# Agent

Use the Agent for flexible foreground operational work, and understand the project setup behind what it can do.

The Agent is the flexible foreground work surface in Lleverage. Use it when the work requires interpretation, investigation, files, connected systems or a sequence that can change depending on what is found.

For example, an operator might ask it to investigate why an order is blocked, prepare supplier follow-up for missing delivery dates, check a set of certificates or explain the current production position using live project context.

The Agent can be used in the Lleverage web app and, when configured, from Slack or Microsoft Teams.

![A live Agent session working through weekly production status with project context and progress visible.](https://3370306531-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FlgSglJDQZxDZdn20CUcl%2Fuploads%2FTgeZREtKbQ06bIHzNCz5%2Fagent-session.png?alt=media)

## What sits behind an Agent

What an Agent can do is shaped by the Project around it:

* **Project description and Project settings** shape general behaviour;
* **Skills** encode recurring methods;
* **Connections** provide authorised access to business systems;
* **Files, Tables, Knowledge and Memory** provide different kinds of context;
* **Workflows** can handle defined parts of the work;
* **Members and Channels** determine who can use it and where.

If you are responsible for creating or changing this setup, go to [Build and improve → Build an Agent](/build-and-improve/overview.md). The builder path covers Project-level instructions, Skills, native and custom Connections, context, testing and operator rollout.

## The Agent is a work surface, not only a chat box

A session keeps the conversation together with evidence around the work. Alongside messages, the session workspace exposes **Progress, Files, Context and Memory**.

* **Progress** — what is happening in a longer task and whether it has completed;
* **Files** — files associated with the task;
* **Context** — tools and referenced material involved;
* **Memory** — remembered project context available beyond the current conversation.

See [Session workspace](/run-operations/agent/session-workspace.md) for the interface walkthrough and current caveats around using those rail surfaces as evidence.

## When the Agent is a good fit

Use the Agent when the route should remain flexible, for example:

* messy emails, files or documents;
* investigation across several sources;
* drafting or recommendation work;
* multi-step work where the next step depends on evidence;
* work that has not yet become a stable repeatable background path.

When a part of the operation becomes sufficiently defined and repeatability matters, it can move into a Workflow without making the whole process non-Agentic.

## When the Agent needs you

The Agent can ask for information or approval inside its session. Respond there when the current Agent task is waiting on you.

That is separate from [Requests](/run-operations/request-inbox.md), which are Workflow-created assigned human handoffs. Agent questions do not automatically become Request Inbox items.

## What good use looks like

Start with a recognisable operational outcome and relevant constraints. Let the Agent work, then inspect the evidence and actual result rather than judging only whether the final response sounds plausible.

If the same correction or instruction is needed repeatedly, improve the setup: usually a Skill, the Project description, a maintained data source, Knowledge source, Connection or Workflow.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.lleverage.ai/run-operations/agent.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.