docs: Cookbook documentation in English updated (#257)

Author: k-farruh

cookbook/en/advanced_deployment.md

@@ -644,9 +644,9 @@ if __name__ == "__main__":
 ```
 
 **Key Points**:
-- Automatically builds and packages project as wheel file
+- Automatically builds and packages the project as a wheel file
 - Uploads artifacts to OSS
-- Creates and manages runtime in AgentRun service
+- Creates and manages runtime in the AgentRun service
 - Automatically creates public access endpoints
 - Supports updating existing deployments (via `agentrun_id` parameter)
 

cookbook/en/call.md

@@ -59,7 +59,7 @@ The following example shows how to reuse `session_id` across multiple requests t
 1. First call: `"My name is Alice."`
 2. Second call: `"What is my name?"`
 
-The SSE stream will mention “Alice”, confirming that session state is in effect.
+The SSE stream will mention “Alice”, confirming that the session state is in effect.
 
 ```python
 session_id = "123456"
@@ -158,5 +158,5 @@ The endpoint reuses the Responses API schema. The `test_openai_compatible_mode`
 - **Parsing errors**: SSE frames must be processed line by line; tolerate keep-alive blank lines or partial JSON chunks.
 - **Context not retained**: Confirm every request includes the same `session_id`, and that the `state/session` services are properly started within `init_func`.
 
-With the examples above you can quickly validate and consume a deployed Agent service.
+With the examples above, you can quickly validate and consume a deployed Agent service.
 

cookbook/en/concept.md

@@ -118,7 +118,7 @@ flowchart LR
 
     class Tools,Service,Sandbox,Adapter,Agent,AgentAPP,Deployer big
     class RT,ST,PT,MS,SS,STS,SBS,BS,FS,GS,CSB,MSB,ETC,TAD,MAD,SAD,STAD,SBAD,AG,AG_NOTE,RA,CT,KD,DP,LD small
-```
+  ```
 
 - **Agent**: The core AI component that processes requests and generates responses; in the runtime we recommend building agents with the AgentScope framework.
 - **AgentApp**: Serves as the application entry point. It exposes APIs, registers routes, loads configurations, and delegates incoming requests to the Runner for execution.
@@ -183,4 +183,4 @@ The runtime offers two patterns for integrating tools:
 
 #### 7. Adapter
 
-`Adapter` is categorized based on different Agent frameworks and includes tool adapters, memory adapters, session adapters, messages protocol adapters, and more.
+`Adapter` is categorized based on different Agent frameworks and includes tool adapters, memory adapters, session adapters, message protocol adapters, and more.

cookbook/en/deployment.md

@@ -29,7 +29,7 @@ Most deployments follow these stages:
 
 ### Service
 
-The `Service` chapter explains the built-in session history, memory, sandbox, and state services plus the shared lifecycle interface. It helps you pick the right implementations (in-memory, Redis, Tablestore, and more) and shows how to manage them via `start()`, `stop()`, and `health()` so your deployment has a reliable backbone. See {doc}`service/service`.
+The `Service` chapter explains the built-in session history, memory, sandbox, and state services, as well as the shared lifecycle interface. It helps you pick the right implementations (in-memory, Redis, Tablestore, and more) and shows how to manage them via `start()`, `stop()`, and `health()` so your deployment has a reliable backbone. See {doc}`service/service`.
 
 ### Simple Deployment
 
@@ -47,7 +47,7 @@ When you need stronger availability or observability guarantees, jump to the adv
 
 - Running multi-service topologies via Docker Compose or Kubernetes.
 - Configuring multi-region/multi-model redundancy, canary releases, and autoscaling.
-- Integrating centralized logging, tracing, and alerting systems.
+- Integrating centralised logging, tracing, and alerting systems.
 
 This is aimed at production scenarios or multi-team collaboration. See {doc}`advanced_deployment` for details.
 
@@ -59,7 +59,7 @@ The reference sample demonstrates a full deployment that bundles sandbox service
 - Building the AgentApp
 - Starting all services
 
-Review {doc}`react_agent` if you want to revisit every deployment step end to end.
+Review {doc}`react_agent` if you want to revisit every deployment step end-to-end.
 
 ## Next Steps
 

cookbook/en/environment_manager.md

@@ -18,7 +18,7 @@ kernelspec:
 
 `EnvironmentManager` provides lifecycle and access to sandboxed environments and tools through `SandboxService`.
 
-Default behavior wires a `SandboxService` instance which manages environment creation, connection, and release.
+By default, a `SandboxService` instance is wired, which manages environment creation, connection, and release.
 
 ## Basic Usage
 
@@ -45,19 +45,19 @@ In the future, `EnvironmentManager` will support not only `sandbox_service` but
 
 ## Sandbox Service
 
-The **Sandbox Service** is designed to manage and provide access to sandboxed tool execution (see {doc}`sandbox` for details) sandboxes for different users and sessions. Sandboxes are organized by a composite key of session ID and user ID, allowing isolated execution contexts for each user session. The service supports multiple sandbox types and can automatically provision the required sandboxes based on the tools being used.
+The **Sandbox Service** is designed to manage and provide access to sandboxed tool execution (see {doc}`sandbox` for details) sandboxes for different users and sessions. Sandboxes are organised by a composite key of session ID and user ID, allowing isolated execution contexts for each user session. The service supports multiple sandbox types and can automatically provision the required sandboxes based on the tools being used.
 
 ### Service Overview
 
-The Sandbox Service provides a unified interface for sandbox management with support for different sandbox types like code execution, file operations, and other specialized sandboxes. The following is an example to initialize a sandbox service:
+The Sandbox Service provides a unified interface for sandbox management, supporting sandbox types such as code execution, file operations, and other specialised sandboxes. The following is an example to initialise a sandbox service:
 
 ```{code-cell}
 from agentscope_runtime.engine.services.sandbox_service import SandboxService
 
 # Create and start the sandbox service
 sandbox_service = SandboxService()
 
-# Or with remote sandbox service
+# Or with a remote sandbox service
 # sandbox_service = SandboxService(
 #     base_url="http://sandbox-server:8000",
 #     bearer_token="your-auth-token"
@@ -68,7 +68,7 @@ sandbox_service = SandboxService()
 
 #### Connecting to Sandboxes
 
-The `connect` method allows you to connect to sandbox sandboxes for a specific user session:
+The `connect` method allows you to connect to the sandbox for a specific user session:
 
 ```{code-cell}
 # Connect with specific sandbox types

cookbook/en/intro.md

@@ -36,7 +36,7 @@ For more detailed change descriptions and the migration guide, please refer to:
 
 ## What is AgentScope Runtime?
 
-**AgentScope Runtime** is a full-stack agent runtime that tackles two core challenges: **efficient agent deployment** and **secure sandbox execution**. It ships with foundational services such as short- and long-term memory plus agent state persistence, along with hardened sandbox infrastructure. Whether you need to orchestrate production-grade agents or guarantee safe tool interactions, AgentScope Runtime provides developer-friendly workflows with complete observability.
+**AgentScope Runtime** is a full-stack agent runtime that tackles two core challenges: **efficient agent deployment** and **secure sandbox execution**. It ships with foundational services such as short- and long-term memory plus agent-state persistence, along with hardened sandbox infrastructure. Whether you need to orchestrate production-grade agents or guarantee safe tool interactions, AgentScope Runtime provides developer-friendly workflows with complete observability.
 
 In V1.0, these services are exposed via an **adapter pattern**, enabling seamless integration with the native modules of different agent frameworks while preserving their native interfaces and behaviors, ensuring both compatibility and flexibility.
 
@@ -46,7 +46,7 @@ This cookbook walks you through building service-ready agent applications with *
 
 **⚙️ Agent Deployment Runtime (Engine)**
 
-Provides `AgentApp` as the main entry point for agent applications, along with production‑grade infrastructure for deploying, managing, and traing agents. It also includes built‑in services such as session history, long‑term memory, and agent state management.
+Provides `AgentApp` as the main entry point for agent applications, along with production‑grade infrastructure for deploying, managing, and training agents. It also includes built‑in services such as session history, long‑term memory, and agent state management.
 
 **🔒 Sandbox Execution Runtime (Sandbox)**
 

cookbook/en/protocol.md

@@ -105,7 +105,7 @@ class Tool(BaseModel):
 ```{code-cell}
 class FunctionCall(BaseModel):
     """
-    Model class for assistant prompt message tool call function.
+    Model class for the assistant prompt message tool call function.
     """
 
     call_id: Optional[str] = None
@@ -128,7 +128,7 @@ class FunctionCall(BaseModel):
 ```{code-cell}
 class FunctionCallOutput(BaseModel):
     """
-    Model class for assistant prompt message tool call function.
+    Model class for the assistant prompt message tool call function.
     """
 
     call_id: str
@@ -240,7 +240,7 @@ class Message(Event):
     """The status of the message. in_progress, completed, or incomplete"""
 
     role: Optional[str] = None
-    """The role of the messages author, should be in `user`,`system`,
+    """The role of messages author, should be in `user`,`system`,
     'assistant'."""
 
     content: Optional[
@@ -540,7 +540,7 @@ ContentBuilder supports multiple content types:
 #### Multi-Content Message Building
 
 ```python
-# Create message containing text and image
+# Create a message containing text and image
 message_builder = response_builder.create_message_builder()
 
 # Add text content
@@ -560,7 +560,7 @@ message_builder.complete()
 #### Data Content Building
 
 ```python
-# Create message containing structured data
+# Create a message containing structured data
 data_builder = message_builder.create_content_builder("data", index=0)
 
 # Set data content
@@ -602,7 +602,7 @@ from agentscope_runtime.engine.deployers.adapter.responses import ResponseAPIDef
 app = AgentApp(agent=agent)
 
 # AgentApp automatically registers built-in adapters
-# You can also customize via protocol_adapters parameter
+# You can also customize via the protocol_adapters parameter
 app = AgentApp(
     agent=agent,
     protocol_adapters=[

cookbook/en/quickstart.md

@@ -67,7 +67,7 @@ print("✅ Dependencies imported successfully")
 
 ### Step 2: Create the Agent App
 
-`AgentApp` is the lifecycle and request hub of the agent application. All initialization, query handling, and shutdown routines are registered on it:
+`AgentApp` is the lifecycle and request hub of the agent application. All initialisation, query handling, and shutdown routines are registered on it:
 
 ```{code-cell}
 agent_app = AgentApp(
@@ -170,7 +170,7 @@ async def query_func(
 
 ### Step 5: Run the Agent App
 
-Start the Agent API server. After launch it listens on `http://localhost:8090/process`:
+Start the Agent API server. After launch, it listens on `http://localhost:8090/process`:
 
 ```{code-cell}
 # Start the service (listen on port 8090)
@@ -221,7 +221,7 @@ async def main():
     await agent_app.deploy(LocalDeployManager(host="0.0.0.0", port=8091))
 ```
 
-This spins up the agent server on the specified port so external clients can call it. Beyond the basic HTTP API, you can also interact with the deployment through protocols like A2A, Response API, and Agent API—see {doc}`protocol` for details.
+This spins up the agent server on the specified port so external clients can call it. Beyond the basic HTTP API, you can also interact with the deployment via protocols such as A2A, Response API, and Agent API—see {doc}`protocol` for details.
 
 Clients can even call the service through the OpenAI SDK:
 

cookbook/en/react_agent.md

@@ -17,7 +17,7 @@ kernelspec:
 This tutorial shows how to build and deploy a *Reason + Act (ReAct)* agent with AgentScope Runtime and the [**AgentScope framework**](https://github.com/modelscope/agentscope).
 
 ```{note}
-The ReAct paradigm interleaves reasoning traces with task-specific actions, which is especially effective for tool-use scenarios. Combining AgentScope's `ReActAgent` with AgentScope Runtime gives you both smart decision making and safe tool execution.
+The ReAct paradigm interleaves reasoning traces with task-specific actions, which is especially effective for tool-use scenarios. Combining AgentScope's `ReActAgent` with AgentScope Runtime gives you both smart decision-making and safe tool execution.
 ```
 
 ## Prerequisites
@@ -107,7 +107,7 @@ async def bootstrap_browser_sandbox():
 asyncio.run(bootstrap_browser_sandbox())
 ```
 
-Here `sandbox_types=["browser"]` matches the test suite so a single browser sandbox instance is reused for the same `session_id` / `user_id` pair.
+Here, `sandbox_types=["browser"]` matches the test suite, so a single browser sandbox instance is reused for the same `session_id` / `user_id` pair.
 
 ### Step 3: Build the AgentApp
 
@@ -265,4 +265,4 @@ A successful result should look like “I'm Friday ...”.
 
 ## Summary
 
-Following these steps you now have a ReAct agent with streaming responses, session memory, browser sandbox tooling, and an OpenAI-compatible endpoint. To deploy remotely or extend the toolset, swap out the model, state services, or registered tools as needed.
+Following these steps, you now have a ReAct agent with streaming responses, session memory, browser sandbox tooling, and an OpenAI-compatible endpoint. To deploy remotely or extend the toolset, swap out the model, state services, or registered tools as needed.