Re-add omitted context

Author: mcongrove

content/Agents/creating-agents.md

@@ -4,15 +4,15 @@ In the Agentuity SDK, agents are one of the core primitives. An agent is where y
 
 ## Framework Flexibility
 
-Agentuity is not framework-specific. You can use any inference library or SDK you prefer—Mastra, Vercel AI SDK, or the model providers' own SDKs directly. Agentuity doesn't care what you use.
+Agentuity is not framework-specific. You can use any inference library or SDK you prefer—Mastra, Vercel AI SDK, the model providers' own SDKs directly, or even a coding agent SDK like Agent SDK from Claude Code. Agentuity doesn't care what you use.
 
 What Agentuity provides is a wrapper around your inference code that makes it easy to add observability, evaluations, structured logging, authentication, and a bunch of other things—without requiring you to set up the underlying infrastructure yourself.
 
 ## Agents as Infrastructure
 
 Agents in Agentuity are meant to be infrastructure pieces and wrappers around the inference calls you would make. The SDK provides easy access to logging, observability, OpenTelemetry traces, and cloud services without having to set up a bunch of stuff—Redis, key-value storage, databases, you name it.
 
-Higher-level services like threads are also provided. If there's a conversation going on, Agentuity will automatically group that conversation together in a thread so the back and forth gets lumped all together. That's an important thing to understand about agents—they're the infrastructure wrapper around your inference code that makes it easy to use.
+Higher-level services like threads are also provided. If there's a conversation going on, Agentuity will automatically group that conversation together in a thread so the back and forth gets lumped all together. With agents, you also get observability and sessions built in. That's an important thing to understand about agents—they're the infrastructure wrapper around your inference code that makes it easy to use.
 
 ## Schema-Driven Development
 

content/Frontend/deployment-scenarios.md

@@ -1,6 +1,6 @@
 # Deployment Scenarios
 
-There are times when you have a Greenfield application—it's important to be able to start fresh. You can have frontend, routes, agents, all the cloud stuff, easy to deploy. But also you have existing applications.
+There are times when you have a Greenfield application—it's important to be able to start fresh. Using `agentuity create`, you can have frontend, routes, agents, all the cloud stuff, easy to deploy. But also you have existing applications.
 
 ## Existing Applications
 

content/Frontend/react-hooks.md

@@ -1,6 +1,6 @@
 # React Hooks
 
-Back to end-to-end type safety—what was happening over and over is someone was building an agent and saying "I need APIs" (which Agentuity offers), then "now I need a frontend for this agent or these things." So Agentuity built a client-side frontend package as well.
+Agents, routes, and frontend hooks are the three big building blocks for building agentic software. Back to end-to-end type safety—what was happening over and over is someone was building an agent and saying "I need APIs" (which Agentuity offers), then "now I need a frontend for this agent or these things." So Agentuity built a client-side frontend package as well.
 
 ## The React Package
 

content/Routes/index.md

@@ -1,6 +1,6 @@
 # Routes
 
-Routes and APIs are another core primitive alongside agents. One thing constantly observed is that people want to build agents, but agents don't live in isolation. Agents have to have lots of different things, including some API or route that you call them or invoke them by.
+Routes and APIs are another core primitive alongside agents—one of the three big building blocks for building what we call agentic software. One thing constantly observed is that people want to build agents, but agents don't live in isolation. Agents have to have lots of different things, including some API or route that you call them or invoke them by.
 
 ## Built on Hono
 

content/Services/Observability/sessions-debugging.md

@@ -18,6 +18,8 @@ Remember—this is an agent-native platform, so agents should have access to the
 
 On deployments, you can see how your deployments are doing. You can view that whole pane of glass by deployment—this deployment was just done, there was a deployment yesterday, let me compare the two. You can see the artifacts, how it's doing, the logs.
 
+On the observability side, you get things like threads, sessions (OpenTelemetry sessions), evals, logs, and analytics—all within a project.
+
 ## CLI Integration
 
 On most screens there's a terminal icon showing how to interact with everything on screen via the CLI. This is orthogonal—it's a cloud platform with a web app to view things, but you also have a CLI to programmatically access everything. An agent can use the CLI for debugging or setting things up programmatically, which is really useful.