Rename koog-ktor-plugin module to koog-ktor and refactor related build files, tests, and code for consistency. Rewrite AI agent functions with improved parameterization and modularity.

Author: nomisRev

build.gradle.kts

@@ -204,7 +204,7 @@ dependencies {
     dokka(project(":prompt:prompt-tokenizer"))
     dokka(project(":prompt:prompt-xml"))
     dokka(project(":koog-spring-boot-starter"))
-    dokka(project(":koog-ktor-plugin"))
+    "dokka"(project(":koog-ktor"))
     dokka(project(":rag:rag-base"))
     dokka(project(":rag:vector-storage"))
 }

examples/build.gradle.kts

@@ -33,8 +33,7 @@ dependencies {
     api(project(":prompt:prompt-executor:prompt-executor-clients:prompt-executor-bedrock-client"))
     api(project(":prompt:prompt-executor:prompt-executor-llms"))
     api(project(":prompt:prompt-executor:prompt-executor-llms-all"))
-
-    api(project(":koog-ktor-plugin"))
+    api(project(":koog-ktor"))
 
     api(libs.kotlinx.datetime)
 

examples/src/main/kotlin/ai/koog/agents/example/ktor/KtorIntegrationExample.kt

@@ -67,7 +67,7 @@ fun Application.configureKoog() {
             fallback { }
         }
 
-        agent {
+        agentConfig {
             mcp {
                 sse("put some url here...")
             }
@@ -111,6 +111,7 @@ private fun Route.agenticRoutes() {
 
         if (isHarmful) {
             call.respond(HttpStatusCode.BadRequest, "Harmful content detected")
+            return@get
         }
 
         val updatedRequest = llm().execute(prompt("id") {

koog-agents/build.gradle.kts

@@ -12,7 +12,7 @@ val excluded = setOf(
     ":examples",
     ":integration-tests",
     ":koog-spring-boot-starter",
-    ":koog-ktor-plugin",
+    ":koog-ktor",
     project.path, // the current project should not depend on itself
 )
 

koog-ktor-plugin/Module.md renamed to koog-ktor/Module.md

@@ -1,16 +1,16 @@
-# Module koog-ktor-plugin
+# Module koog-ktor
 
 Ktor server integration for the Koog AI agents framework.
 
 ## Overview
 
-The `koog-ktor-plugin` module provides seamless integration between the Koog AI agents framework and Ktor server applications. It includes:
+The `koog-ktor` module provides seamless integration between the Koog AI agents framework and Ktor server applications.
+It includes:
 
 - A Ktor plugin for easy installation and configuration
 - Support for multiple LLM providers (OpenAI, Anthropic, Google, OpenRouter, Ollama)
 - Agent configuration with tools, features, and prompt customization
 - Extension functions for routes to interact with LLMs and agents
-- Content moderation capabilities
 - JVM-specific support for Model Context Protocol (MCP) integration
 
 ## Using in your project
@@ -19,58 +19,47 @@ Add the dependency to your `build.gradle.kts`:
 
 ```kotlin
 dependencies {
-    implementation("ai.koog:koog-ktor-plugin:$koogVersion")
+    implementation("ai.koog:koog-ktor:$koogVersion")
 }
 ```
 
 ## Basic Usage
 
-Install and configure the plugin in your Ktor application:
+Provide one, or many apikey-s, and the Koog plugin will automatically connect to the provider when needed.
+For additional, or provider-specific settings, See [YAML/CONF Configuration](#yamlconf-configuration) below.
+
+```yaml
+koog:
+  openai.apikey: "$OPENAI_API_KEY:your-openai-api-key"
+  anthropic.apikey: "$ANTHROPIC_API_KEY:your-anthropic-api-key"
+  google.apikey: "$GOOGLE_API_KEY:your-google-api-key"
+  openrouter.apikey: "$OPENROUTER_API_KEY:your-openrouter-api-key"
+  ollama.enabled: "$DEBUG:false"
+```
+
+### Installing and configuring the plugin
+
+The Koog plugin can also be configured by code, and some complex configurations can only be done by code.
+See [Programmatic Configuration](#programmatic-configuration) below.
 
 ```kotlin
 fun Application.module() {
     install(Koog) {
-        // Configure LLM providers
         llm {
-            openAI(apiKey = "your-openai-api-key") {
-                baseUrl = "https://api.openai.com"
-            }
-            
-            // Optional: Configure other providers
+            openAI(apiKey = "your-openai-api-key")
             anthropic(apiKey = "your-anthropic-api-key")
             ollama { baseUrl = "http://localhost:11434" }
             google(apiKey = "your-google-api-key")
             openRouter(apiKey = "your-openrouter-api-key")
         }
-        
-        // Configure agent
-        agent {
-            // Register tools
-            registerTools {
-                // Add tools using reflection
-                tool(::yourToolFunction)
-            }
-            
-            // Configure prompt
-            prompt {
-                system("You are a helpful assistant")
-            }
-            
-            // JVM-specific: Configure MCP integration
-            mcp {
-                sse("your-mcp-server-url")
-            }
-        }
     }
-    
-    // Use in routes
+
     routing {
         route("/ai") {
             post("/chat") {
                 val userInput = call.receive<String>()
-                
-                // Use agent to respond
-                call.agentRespond(userInput)
+                val output = singleRunAgent(userInput)
+                call.respond(HttpStatusCode.OK, output)
             }
         }
     }
@@ -84,19 +73,19 @@ fun Application.module() {
 ```kotlin
 post("/moderated-chat") {
     val userInput = call.receive<String>()
-    
+
     // Moderate content
-    val isHarmful = moderateWithLLM(OpenAIModels.Moderation.Omni) {
-        user(userInput)
-    }.isHarmful
-    
+    val isHarmful = llm().moderate(prompt("id") {
+        user(userRequest)
+    }, OpenAIModels.Moderation.Omni).isHarmful
+
     if (isHarmful) {
         call.respond(HttpStatusCode.BadRequest, "Harmful content detected")
         return@post
     }
-    
-    // Process with agent
-    call.agentRespond(userInput)
+
+    val output = singleRunAgent(userInput)
+    call.respond(HttpStatusCode.OK, output)
 }
 ```
 
@@ -105,14 +94,16 @@ post("/moderated-chat") {
 ```kotlin
 post("/llm-chat") {
     val userInput = call.receive<String>()
-    
-    // Ask LLM directly
-    val response = askLLM(OllamaModels.Meta.LLAMA_3_2) {
-        system("You are a helpful assistant")
-        user(userInput)
-    }.single() as Message.Assistant
-    
-    call.respond(response.content)
+
+    val response = llm().execute(prompt("id") {
+        system(
+            "You are a helpful assistant that can correct user answers. " +
+                    "You will get a user's question and your task is to make it more clear for the further processing."
+        )
+        user(userRequest)
+    }, OllamaModels.Meta.LLAMA_3_2)
+
+    call.respond(HttpStatusCode.OK, response.content)
 }
 ```
 
@@ -121,9 +112,9 @@ post("/llm-chat") {
 ```kotlin
 post("/custom-agent") {
     val userInput = call.receive<String>()
-    
-    // Use custom strategy
-    call.agentRespond(userInput, strategy = reActStrategy())
+
+    val output = aiAgent(reActStrategy(), userInput)
+    call.respond(HttpStatusCode.OK, output)
 }
 ```
 
@@ -145,7 +136,7 @@ llm {
             socketTimeoutMillis = 30000
         }
     }
-    
+
     // Set fallback LLM
     fallback {
         provider = LLMProvider.Ollama
@@ -156,7 +147,8 @@ llm {
 
 #### YAML/CONF Configuration
 
-You can also configure LLM providers using YAML or CONF files. The plugin will automatically read the configuration from the application's configuration file:
+You can also configure LLM providers using YAML or CONF files. The plugin will automatically read the configuration from
+the application's configuration file:
 
 ```yaml
 # application.yaml or application.conf
@@ -168,28 +160,29 @@ koog:
       requestTimeoutMillis: 30000
       connectTimeoutMillis: 10000
       socketTimeoutMillis: 30000
-  
+
   anthropic:
     apikey: "your-anthropic-api-key"
     baseUrl: "https://api.anthropic.com"
     timeout:
       requestTimeoutMillis: 30000
-  
+
   google:
     apikey: "your-google-api-key"
     baseUrl: "https://generativelanguage.googleapis.com"
-  
+
   openrouter:
     apikey: "your-openrouter-api-key"
     baseUrl: "https://openrouter.ai"
-  
+
   ollama:
     baseUrl: "http://localhost:11434"
     timeout:
       requestTimeoutMillis: 60000
 ```
 
-When using configuration files, you can still provide programmatic configuration that will override the settings from the file:
+When using configuration files, you can still provide programmatic configuration that will override the settings from
+the file:
 
 ```kotlin
 install(Koog) {
@@ -198,7 +191,7 @@ install(Koog) {
         // This will override the API key from the configuration file
         openAI(apiKey = System.getenv("OPENAI_API_KEY") ?: "override-from-code")
     }
-    
+
     // Rest of your configuration...
 }
 ```
@@ -211,21 +204,21 @@ Configure agent behavior, tools, and features:
 agent {
     // Set model
     model = OpenAIModels.GPT4.Turbo
-    
+
     // Set max iterations
     maxAgentIterations = 10
-    
+
     // Register tools
     registerTools {
         tool(::searchTool)
         tool(::calculatorTool)
     }
-    
+
     // Configure prompt
     prompt {
         system("You are a helpful assistant specialized in...")
     }
-    
+
     // Install features
     install(OpenTelemetry) {
         // Configure feature
@@ -242,10 +235,10 @@ agent {
     mcp {
         // Use Server-Sent Events
         sse("https://your-mcp-server.com/sse")
-        
+
         // Or use process
         process(yourMcpProcess)
-        
+
         // Or use existing client
         client(yourMcpClient)
     }