docs(blocks): give the Agent intro the weight it deserves

The opening was too flat for the central block. New lead: 'The Agent block is
where a workflow thinks' — model, reasoning, tools, structured output, and the
contrast with data-moving blocks. The agent-vs-Agent-block disambiguation moves
from an info callout into a second paragraph framed as the block's role in
building agents (the unit you build agents from). Dropped the now-unused
Callout import.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

Author: ouiliame

apps/docs/content/docs/en/blocks/agent.mdx

@@ -4,17 +4,14 @@ description: The Agent block runs a model over your inputs to reason, call tools
 pageType: reference
 ---
 
-import { Callout } from 'fumadocs-ui/components/callout'
 import { BlockPreview, WorkflowPreview, CLASSIFY_WORKFLOW } from '@/components/workflow-preview'
 import { FAQ } from '@/components/ui/faq'
 
-The **Agent block** runs a model as one step of a workflow. It reads the values you give it, optionally calls tools and reads files, and returns a result, either plain text or structured JSON, that later blocks read by reference.
+The **Agent block** is where a workflow thinks. It runs a model with the instructions, context, and tools you give it: the model reads the input, reasons about it, decides which tools to call, and returns plain text or structured JSON for the rest of the workflow to build on. Other blocks move and shape data; the Agent block is the step that understands it.
 
-<BlockPreview type="agent" />
+It's also the unit you build agents from. An *agent* — the thing you ship — is a whole workflow that reasons and acts on its own; an *Agent block* is one reasoning step inside it. The simplest agent is a single Agent block with tools. Larger ones wire several together, with conditions, loops, and other blocks around them. See [Building agents](/building-agents).
 
-<Callout type="info">
-An *agent* and an *Agent block* are not the same thing. An agent is a workflow you build and ship that reasons and acts on its own. An Agent block is one reasoning step inside it. A simple agent is a single Agent block; a larger one wires several together with other blocks. See [Building agents](/building-agents).
-</Callout>
+<BlockPreview type="agent" />
 
 ## Configuration