> 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/agent-context.md).

# Context and capabilities

Choose the right source for Agent instructions, methods, operational data, working files, reference knowledge and persistent context.

The Agent can draw on several kinds of instructions, evidence and capability. The important distinctions are:

* **what is available to the current run**;
* **what was actually used for this task**; and
* **which source is authoritative when information conflicts**.

## The main context layers

* **Project description** gives durable Project-level direction about the operational area and general behaviour.
* **Skills** provide reusable methods for recurring work.
* **Files** provide task inputs, evidence and outputs.
* **Integrations** give authorised access to external business systems.
* **Access and approvals** explain what grants authority and where human approval can be required.
* **Knowledge Bases** provide maintained reference material.
* **Tables** provide maintained structured data.
* **Memory** carries useful persistent context across sessions.
* **Workflows** can provide repeatable execution for stable parts of the job.

Lleverage can also provide built-in Skills and general industry/process knowledge. Use that as a starting point, not as authority over customer-specific rules, policies, master data or live system records.

## Scope comes before capability

In a web Agent session, the current Project is explicit.

Slack and Microsoft Teams can also run at organisation scope without one fixed Project. In that case, the run is limited to the acting user's authorised Project scope and Project-specific capabilities retain their provenance.

Pin a channel conversation when the work should stay inside one Project or when several Projects contain overlapping capabilities.

## What the session shows

The session's **Context** area helps you inspect tools and referenced material around the current task.

That view is supporting evidence, not a complete source-of-truth ledger. A tool appearing in Context does not make its data authoritative, and an empty panel should not be treated as proof that no capability contributed.

For important work, verify the operational result against the system, document or maintained source that should own it.

See [Access and approvals](/run-operations/agent/agent-context/access-and-approvals.md) when the question is what the Agent is allowed to read, change or execute.

## Source authority matters

These are different kinds of evidence:

* a current ERP record;
* an approved policy;
* a maintained Project mapping;
* remembered context;
* a file uploaded for this case;
* generic industry guidance;
* an Agent inference.

Do not let them silently become interchangeable.

When the distinction matters, define which source wins and what the Agent should do when sources disagree.

For the session-level interface, see [Session workspace](/run-operations/agent/session-workspace.md). For the broader setup model, see [How Lleverage gets its context](/understand-lleverage/readme/how-lleverage-gets-its-context.md).


---

# 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/agent-context.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.