[pull] main from google:main

README.md

@@ -3,7 +3,7 @@
 [![License](https://img.shields.io/badge/License-Apache_2.0-blue.svg)](LICENSE)
 [![PyPI](https://img.shields.io/pypi/v/google-adk)](https://pypi.org/project/google-adk/)
 [![NPM Version](https://img.shields.io/npm/v/@google/adk)](https://www.npmjs.com/package/@google/adk)
-[![Go Doc](https://img.shields.io/badge/Go%20Package-Doc-blue.svg)](https://pkg.go.dev/google.golang.org/adk)
+[![Go Doc](https://img.shields.io/badge/Go%20Package-Doc-blue.svg)](https://pkg.go.dev/google.golang.org/adk/v2)
 [![Maven Central Java](https://img.shields.io/maven-central/v/com.google.adk/google-adk)](https://search.maven.org/artifact/com.google.adk/google-adk)
 [![Maven Central Kotlin](https://img.shields.io/maven-central/v/com.google.adk/google-adk-kotlin-core)](https://search.maven.org/artifact/com.google.adk/google-adk-kotlin-core)
 

docs/agents/models/litert-lm.md

@@ -1,22 +1,73 @@
 # LiteRT-LM model host for ADK agents
 
 <div class="language-support-tag">
-    <span class="lst-supported">Supported in ADK</span><span class="lst-python">Python v0.1.0</span>
+    <span class="lst-supported">Supported in ADK</span><span class="lst-python">Python v0.1.0</span><span class="lst-kotlin">Kotlin v0.4.0</span>
 </div>
 
-[LiteRT-LM](https://github.com/google-ai-edge/LiteRT-LM) is a C++ library to
-efficiently run language models across edge platforms.
-On desktop (Linux, macOS, and Windows), ADK integrates with LiteRT-LM-hosted
-models through the LiteRT-LM server launched by the LiteRT-LM CLI `lit`.
+You can use the
+[LiteRT-LM](https://github.com/google-ai-edge/LiteRT-LM) 
+library to efficiently run language models locally on various compute devices
+without requiring specialized processors such as graphics processing units 
+(GPU) or tensor processing units (TPU). LiteRT-LM supports many models,
+including Google Gemma models as well as third-party models. This guide 
+provides instructions for setting up LiteRT-LM with ADK for the following
+languages:
 
-## Get started
+- [Python](#python) 
+- [Kotlin](#kotlin)
 
-LiteRT-LM works with the `Gemini` class. You only have to set the `base_url` and
-`model` parameters.
+## Python
+
+These instructions describe how to use LiteRT-LM server with ADK in Python
+with a Gemma open weights model, including using LiteRT-LM 's local hosting
+model server `lit`.
+
+### Install resources
+
+You need to download a model to use with LiteRT-LM, and the `lit` CLI tool
+to help you find a model download it.
+
+#### Install `lit` CLI tool
+
+Download and install the `lit` CLI tool by following these
+[instructions](https://github.com/google-ai-edge/LiteRT-LM?tab=readme-ov-file#desktop-cli-lit)
+in the LiteRT-LM GitHub repository.
+
+#### Download a model
+
+Before you start the server, you need to download a model. You'll need a
+_Hugging Face_ user access token to download a LiteRT-LM model using `lit`. You
+can get a token for your _Hugging Face_ account
+[here](https://huggingface.co/settings/tokens).
+
+To see a list of models available for download, use the `lit list` command:
+
+```bash
+lit list --show_all
+```
+
+Download a model using the `lit pull` command:
+
+```bash
+export HUGGING_FACE_HUB_TOKEN="**your Hugging Face token**"
+lit pull gemma3n-e2b
+```
+
+### Configure your agent
+
+Configure your agent to connect to LiteRT-LM and a hosted model.
+When running Gemma models with LiteRT-LM, you configure a `Gemini`
+model class with the model identifier and local network address.
+
+To use LiteRT-LM with ADK and a Gemma model:
 
 1.  Set `base_url` to the LiteRT-LM server URL, for example: `localhost:8001`.
 2.  Set `model` to the LiteRT-LM model name, for example: `gemma3n-e2b`.
 
+The following example code shows how to configure an agent to
+connect to the locally hosted LiteRT-LM instance serving the Gemma
+model configuration described above:
+
 ```py
 from google.adk.agents import Agent
 from google.adk.models import Gemini
@@ -47,55 +98,167 @@ Then run the agent as usual:
 adk web
 ```
 
-## Running the LiteRT-LM Server
+### Running the LiteRT-LM server
 
 The LiteRT-LM server is a separate process that serves LiteRT-LM models. It is
 started by the LiteRT-LM CLI tool `lit`.
 
-### Download the `lit` CLI tool
+#### Run the server
 
-Download the `lit` CLI tool by following these
-[instructions](https://github.com/google-ai-edge/LiteRT-LM?tab=readme-ov-file#desktop-cli-lit)
-in the LiteRT-LM GitHub repository.
+After downloading a model, start the LiteRT-LM server locally by running the
+following command:
+
+```bash
+lit serve --port 8001
+```
 
-### Download a model
+!!! tip "Local Server Port Number"
 
-Before you start the server, you need to download a model. You'll need a
-*Hugging Face* user access token to download a LiteRT-LM model using `lit`. You
-can get a token for your *Hugging Face* account
-[here](https://huggingface.co/settings/tokens).
+    You may choose any port number for the LiteRT-LM server as long as it matches the `base_url` you set in the `Gemini` class in your agent code.
 
-To see a list of models available for download, use the `lit list` command:
+#### Debugging
+
+To see incoming requests to the LiteRT-LM server and the exact input sent to the
+model, use the `--verbose` flag:
 
 ```bash
-lit list --show_all
+lit serve --port 8001 --verbose
 ```
 
-Download a model using the `lit pull` command:
+## Kotlin
+
+These instructions describe how to use LiteRT-LM with ADK in Kotlin using
+the `com.google.adk.kt.litertlm` package.
+
+### Install resources
+
+You need to download a model to use with LiteRT-LM, and the `litert-lm` CLI tool
+to help you find a model download it.
+
+#### Install LiteRT-LM CLI
+
+Prerequisites: Python 3.10 or higher
+
+To install the CLI, run:
 
 ```bash
-export HUGGING_FACE_HUB_TOKEN="**your Hugging Face token**"
-lit pull gemma3n-e2b
+pip install --upgrade litert-lm
 ```
 
-### Run the server
+For additional installation methods, such as using uv, see
+[LiteRT-LM CLI Installation Guide](https://developers.google.com/edge/litert-lm/cli/installation).
 
-After downloading a model, start the LiteRT-LM server locally by running the
-following command:
+#### Download a model
+
+Download a model compatible with LiteRT-LM to use the `litert-lm` CLI tool. 
+Use `litert-lm` to download models directly from Hugging Face:
 
 ```bash
-lit serve --port 8001
+litert-lm import \
+  --from-huggingface-repo litert-community/gemma-4-E2B-it-litert-lm \
+  gemma-4-E2B-it.litertlm
 ```
 
-!!! tip "Local Server Port Number"
+Once downloaded, the model is stored locally at:
 
-    You may choose any port number for the LiteRT-LM server as long as it matches the `base_url` you set in the `Gemini` class in your agent code.
+```
+~/.litert-lm/models/gemma-4-E2B-it.litertlm/model.litertlm
+```
 
-### Debugging
+For more details about `litert-lm`, refer to the
+[LiteRT-LM CLI Usage Guide](https://developers.google.com/edge/litert-lm/cli/usage).
 
-To see incoming requests to the LiteRT-LM server and the exact input sent to the
-model, use the `--verbose` flag:
+### Add dependencies
+
+ADK Kotlin works with LiteRT-LM through an adapter package,
+`com.google.adk:google-adk-kotlin-litertlm`.
+
+In your `build.gradle.kts`, add `com.google.adk:google-adk-kotlin-litertlm` and
+`com.google.ai.edge.litertlm:litertlm-jvm` to your dependencies:
+
+```kt
+repositories {
+    mavenCentral()
+    google()
+}
+
+dependencies {
+    implementation("com.google.adk:google-adk-kotlin-core:0.4.0")
+    implementation("com.google.adk:google-adk-kotlin-litertlm:0.4.0")
+    implementation("com.google.ai.edge.litertlm:litertlm-jvm:0.13.1")
+    // other dependencies...
+}
+```
+
+### Configure agent model
+
+Run a local model for your agent with LiteRT-LM by configuring a
+`LiteRtLmModel` object as part of your `LlmAgent` object. If you do not
+already have a ADK Kotlin project, follow the 
+[Kotlin Quickstart for ADK](/get-started/kotlin/) 
+getting started guide. The following code example shows you how to
+configure an `LlmAgent`, and set the `model` parameter to a `LiteRtLmModel`:
+
+```kt
+ object HelloTimeAgent {
+
+    // Get model path from environment variable.
+    private val modelPath: String by lazy {
+        System.getenv("LITERT_LM_MODEL_PATH")
+            ?: throw IllegalStateException(
+                "LITERT_LM_MODEL_PATH environment variable must be set pointing to a .litertlm file."
+            )
+    }
+
+    @JvmField
+    val rootAgent =
+        LlmAgent(
+            name = "hello_time_agent",
+            description = "Tells the current time in a specified city.",
+            model =
+                LiteRtLmModel.create(
+                    EngineConfig(modelPath = modelPath, backend = Backend.CPU())
+                ),
+            instruction =
+                Instruction(
+                    "You are a helpful assistant that tells the current time in a city. " +
+                        "Use the 'getCurrentTime' tool for this purpose."
+                ),
+            tools = TimeService().generatedTools(),
+        )
+}
+```
+
+In this example, the path to the LiteRT-LM model file is read from the
+environment variable `LITERT_LM_MODEL_PATH`. The model will be run on the CPU.
+You can run the model on a GPU by setting `backend = Backend.GPU()`.
+
+When you run the agent, set `LITERT_LM_MODEL_PATH` to the location of the model
+file, for example: `~/.litert-lm/models/gemma-4-E2B-it.litertlm/model.litertlm`.
+
+### Run your agent
+
+If you followed the [Kotlin Quickstart for ADK](/get-started/kotlin/)
+with the above modifications, you can run your ADK agent using the command-line 
+REPL with the environment variable `LITERT_LM_MODEL_PATH` set to the path of
+the model file:
 
 ```bash
-lit serve --port 8001 --verbose
+LITERT_LM_MODEL_PATH=~/.litert-lm/models/gemma-4-E2B-it.litertlm/model.litertlm ./gradlew run
+```
+
+Example interaction:
+
+```
+Agent hello_time_agent is ready. Type 'exit' to quit.
+
+You > what's your name?
+
+hello_time_agent > I am Gemma 4, a Large Language Model developed by Google DeepMind.
+
+You > what time is it in paris?
+
+hello_time_agent > calls tool: getCurrentTime
+
+hello_time_agent > The time in Paris is 10:30 am.
 ```

docs/api-reference/index.md

@@ -2,6 +2,16 @@
 
 The Agent Development Kit (ADK) provides comprehensive API references across supported languages, allowing you to dive deep into all available classes, methods, and functionalities.
 
+{{#
+  CONTRIBUTORS: Keep all API reference links below as full `https://` absolute URLs.
+  These API docs (python/, typescript/, java/, kotlin/, cli/, agentconfig/, rest/)
+  are externally generated HTML, not MkDocs Markdown pages. If written as relative
+  or root-relative paths, such as `python/index.html` or `/api-reference/python/`,
+  the mkdocs-llmstxt plugin rewrites them to non-existent `.md` URLs, such as
+  `https://adk.dev/api-reference/python/index.md`, producing broken 404 links in
+  llms.txt / llms-full.txt. See #1716 and #1717 for history.
+#}}
+
 <div class="grid cards" markdown>
 
 -   :fontawesome-brands-python:{ .lg .middle } **Python API Reference**
@@ -10,9 +20,6 @@ The Agent Development Kit (ADK) provides comprehensive API references across sup
     Explore the complete API documentation for the Python Agent Development Kit. Discover detailed information on all modules, classes, functions, and examples to build sophisticated AI agents with Python.
 
     [:octicons-arrow-right-24: View Python API Docs](https://adk.dev/api-reference/python/) <br>
-    <!-- Assuming your Python API docs are in a 'python' subdirectory -->
-    <!-- Or link to an external ReadTheDocs, etc. -->
-    <!-- [:octicons-arrow-right-24: View Python API Docs](python/index.html) -->
 
 <!-- This comment forces a block separation -->
 
@@ -22,8 +29,6 @@ The Agent Development Kit (ADK) provides comprehensive API references across sup
     Access the complete API documentation for the TypeScript Agent Development Kit. Find detailed information on all packages, classes, and methods to build powerful and flexible AI agents with TypeScript.
 
     [:octicons-arrow-right-24: View Typescript API Docs](https://adk.dev/api-reference/typescript/) <br>
-    <!-- Assuming your Typescript API docs are in a 'typescript' subdirectory -->
-    <!-- [:octicons-arrow-right-24: View Typescript API Docs](typescript/index.html) -->
 
 <!-- This comment forces a block separation -->
 
@@ -32,7 +37,8 @@ The Agent Development Kit (ADK) provides comprehensive API references across sup
     ---
     Explore the complete API documentation for the Go Agent Development Kit. Discover detailed information on all modules, classes, and functions to build sophisticated AI agents with Go.
 
-    [:octicons-arrow-right-24: View Go API Docs](https://pkg.go.dev/google.golang.org/adk) <br>
+    [:octicons-arrow-right-24: View Go v2.x API Docs](https://pkg.go.dev/google.golang.org/adk/v2) <br>
+    [:octicons-arrow-right-24: View Go v1.x API Docs](https://pkg.go.dev/google.golang.org/adk) <br>
 
 <!-- This comment forces a block separation -->
 
@@ -42,9 +48,6 @@ The Agent Development Kit (ADK) provides comprehensive API references across sup
     Access the comprehensive Javadoc for the Java Agent Development Kit. This reference provides detailed specifications for all packages, classes, interfaces, and methods, enabling you to develop robust AI agents using Java.
 
     [:octicons-arrow-right-24: View Java API Docs](https://adk.dev/api-reference/java/) <br>
-    <!-- Assuming your Java API docs (Javadocs) are in a 'java' subdirectory -->
-    <!-- Or link to an external Javadoc hosting site -->
-    <!-- [:octicons-arrow-right-24: View Java API Docs](java/index.html) -->
 
 <!-- This comment forces a block separation -->
 

docs/integrations/mimir.md

@@ -145,4 +145,4 @@ session = await runner.session_service.create_session(
 - [adk-mimir-memory on GitHub](https://github.com/Perseus-Computing-LLC/adk-mimir-memory)
 - [adk-mimir-memory on PyPI](https://pypi.org/project/adk-mimir-memory/)
 - [Mimir (backing service)](https://github.com/Perseus-Computing-LLC/mimir)
-- [Perseus (context engine)](https://github.com/Perseus-Computing-LLC/perseus)
+- [Perseus Context integration](/integrations/perseus/)

docs/runtime/runconfig.md

@@ -324,5 +324,6 @@ your language:
 
 - [Python API reference](../api-reference/python/google-adk.html#google.adk.agents.RunConfig)
 - [TypeScript API reference](../api-reference/typescript/interfaces/RunConfig.html)
-- [Go API reference](https://pkg.go.dev/google.golang.org/adk/agent#RunConfig)
+- [Go API reference](https://pkg.go.dev/google.golang.org/adk/v2/agent#RunConfig)
 - [Java API reference](../api-reference/java/com/google/adk/agents/RunConfig.html)
+- [Kotlin API reference](../api-reference/kotlin/google-adk-kotlin-core/com.google.adk.kt.agents/-run-config/index.html)

docs/skills/index.md

@@ -287,11 +287,11 @@ You can define Skills within the code of your agent, as shown below.
     ```go
     import (
         "os"
-    
+
         "google.golang.org/adk/tool/skilltoolset/skill"
         "google.golang.org/adk/tool/skilltoolset"
     )
-    
+
     // ...
 
     source := skill.NewFileSystemSource(os.DirFS("./skills"))
@@ -301,7 +301,7 @@ You can define Skills within the code of your agent, as shown below.
     //   source, _, err = skill.WithFrontmatterPreloadSource(ctx, source)
     //   source, _, err = skill.WithCompletePreloadSource(ctx, source)
     // For more information about these and other wrappers, see
-    // https://pkg.go.dev/google.golang.org/adk/tool/skilltoolset/skill#Source.
+    // https://pkg.go.dev/google.golang.org/adk/v2/tool/skilltoolset/skill#Source.
 
     skillToolset, err := skilltoolset.New(ctx, skilltoolset.Config{
         Source: source,