Skip to content

循环智能体

ADK 中支持Python v0.1.0Typescript v0.2.0Go v0.1.0Java v0.2.0

The LoopAgent is a workflow agent that executes its sub-agents in a loop (i.e. iteratively). It repeatedly runs a sequence of agents for a specified number of iterations or until a termination condition is met.

当您的工作流程涉及重复或迭代优化时 (例如,修订代码),请使用 LoopAgent。

示例

  • You want to build an agent that can generate images of food, but sometimes when you want to generate a specific number of items (e.g. 5 bananas), it generates a different number of those items in the image, such as an image of 7 bananas. You have two tools: Generate Image, Count Food Items. Because you want to keep generating images until it either correctly generates the specified number of items, or after a certain number of iterations, you should build your agent using a LoopAgent.

与其他 工作流智能体 一样,LoopAgent 不由 LLM 驱动,因此在执行方式上是确定性的。也就是说,工作流智能体只关心其执行 (例如,在循环中),而不关心其内部逻辑;工作流智能体的工具或子智能体可能利用也可能不利用 LLM。

工作原理

When the LoopAgent's Run Async method is called, it performs the following actions:

  1. 子智能体执行: 它按顺序迭代子智能体列表。对于每个子智能体,它调用该智能体的 Run Async 方法。

  2. 终止检查:

    至关重要的是,LoopAgent 本身不本质上决定何时停止循环。您必须实现终止机制以防止无限循环。常见策略包括:

    • 最大迭代次数: 在 LoopAgent 中设置最大迭代次数。循环将在该多次迭代后终止。
    • 来自子智能体的升级: 设计一个或多个子智能体来评估条件 (例如,"文档质量是否足够好?"、"是否已达成共识?")。如果满足条件,子智能体可以发出终止信号 (例如,通过引发自定义事件,在共享上下文中设置标志,或返回特定值)。

Loop Agent

完整示例:迭代文档改进

Imagine a scenario where you want to iteratively improve a document:

  • Writer Agent: 生成或完善主题草稿的 LlmAgent。

  • Critic Agent: 批评草稿并确定需要改进的区域的 LlmAgent。

    LoopAgent(sub_agents=[WriterAgent, CriticAgent], max_iterations=5)

In this setup, the LoopAgent would manage the iterative process. The CriticAgent could be designed to return a "STOP" signal when the document reaches a satisfactory quality level, preventing further iterations. Alternatively, the max iterations parameter could be used to limit the process to a fixed number of cycles, or external logic could be implemented to make stop decisions. The loop would run at most five times, ensuring the iterative refinement doesn't continue indefinitely.

完整代码 
# Part of agent.py --> Follow https://google.github.io/adk-docs/get-started/quickstart/ to learn the setup

import asyncio
import os
from google.adk.agents import LoopAgent, LlmAgent, BaseAgent, SequentialAgent
from google.genai import types
from google.adk.runners import InMemoryRunner
from google.adk.agents.invocation_context import InvocationContext
from google.adk.tools.tool_context import ToolContext
from typing import AsyncGenerator, Optional
from google.adk.events import Event, EventActions

# --- Constants ---
APP_NAME = "doc_writing_app_v3" # New App Name
USER_ID = "dev_user_01"
SESSION_ID_BASE = "loop_exit_tool_session" # New Base Session ID
GEMINI_MODEL = "gemini-2.0-flash"
STATE_INITIAL_TOPIC = "initial_topic"

# --- State Keys ---
STATE_CURRENT_DOC = "current_document"
STATE_CRITICISM = "criticism"
# Define the exact phrase the Critic should use to signal completion
COMPLETION_PHRASE = "No major issues found."

# --- Tool Definition ---
def exit_loop(tool_context: ToolContext):
  """Call this function ONLY when the critique indicates no further changes are needed, signaling the iterative process should end."""
  print(f"  [Tool Call] exit_loop triggered by {tool_context.agent_name}")
  tool_context.actions.escalate = True
  # Return empty dict as tools should typically return JSON-serializable output
  return {}

# --- Agent Definitions ---

# STEP 1: Initial Writer Agent (Runs ONCE at the beginning)
initial_writer_agent = LlmAgent(
    name="InitialWriterAgent",
    model=GEMINI_MODEL,
    include_contents='none',
    # MODIFIED Instruction: Ask for a slightly more developed start
    instruction=f"""You are a Creative Writing Assistant tasked with starting a story.
    Write the *first draft* of a short story (aim for 2-4 sentences).
    Base the content *only* on the topic provided below. Try to introduce a specific element (like a character, a setting detail, or a starting action) to make it engaging.
    Topic: {{initial_topic}}

    Output *only* the story/document text. Do not add introductions or explanations.
""",
    description="Writes the initial document draft based on the topic, aiming for some initial substance.",
    output_key=STATE_CURRENT_DOC
)

# STEP 2a: Critic Agent (Inside the Refinement Loop)
critic_agent_in_loop = LlmAgent(
    name="CriticAgent",
    model=GEMINI_MODEL,
    include_contents='none',
    # MODIFIED Instruction: More nuanced completion criteria, look for clear improvement paths.
    instruction=f"""You are a Constructive Critic AI reviewing a short document draft (typically 2-6 sentences). Your goal is balanced feedback.

    **Document to Review:**
    ```
    {{current_document}}
    ```

    **Task:**
    Review the document for clarity, engagement, and basic coherence according to the initial topic (if known).

    IF you identify 1-2 *clear and actionable* ways the document could be improved to better capture the topic or enhance reader engagement (e.g., "Needs a stronger opening sentence", "Clarify the character's goal"):
    Provide these specific suggestions concisely. Output *only* the critique text.

    ELSE IF the document is coherent, addresses the topic adequately for its length, and has no glaring errors or obvious omissions:
    Respond *exactly* with the phrase "{COMPLETION_PHRASE}" and nothing else. It doesn't need to be perfect, just functionally complete for this stage. Avoid suggesting purely subjective stylistic preferences if the core is sound.

    Do not add explanations. Output only the critique OR the exact completion phrase.
""",
    description="Reviews the current draft, providing critique if clear improvements are needed, otherwise signals completion.",
    output_key=STATE_CRITICISM
)


# STEP 2b: Refiner/Exiter Agent (Inside the Refinement Loop)
refiner_agent_in_loop = LlmAgent(
    name="RefinerAgent",
    model=GEMINI_MODEL,
    # Relies solely on state via placeholders
    include_contents='none',
    instruction=f"""You are a Creative Writing Assistant refining a document based on feedback OR exiting the process.
    **Current Document:**
    ```
    {{current_document}}
    ```
    **Critique/Suggestions:**
    {{criticism}}

    **Task:**
    Analyze the 'Critique/Suggestions'.
    IF the critique is *exactly* "{COMPLETION_PHRASE}":
    You MUST call the 'exit_loop' function. Do not output any text.
    ELSE (the critique contains actionable feedback):
    Carefully apply the suggestions to improve the 'Current Document'. Output *only* the refined document text.

    Do not add explanations. Either output the refined document OR call the exit_loop function.
""",
    description="Refines the document based on critique, or calls exit_loop if critique indicates completion.",
    tools=[exit_loop], # Provide the exit_loop tool
    output_key=STATE_CURRENT_DOC # Overwrites state['current_document'] with the refined version
)


# STEP 2: Refinement Loop Agent
refinement_loop = LoopAgent(
    name="RefinementLoop",
    # Agent order is crucial: Critique first, then Refine/Exit
    sub_agents=[
        critic_agent_in_loop,
        refiner_agent_in_loop,
    ],
    max_iterations=5 # Limit loops
)

# STEP 3: Overall Sequential Pipeline
# For ADK tools compatibility, the root agent must be named `root_agent`
root_agent = SequentialAgent(
    name="IterativeWritingPipeline",
    sub_agents=[
        initial_writer_agent, # Run first to create initial doc
        refinement_loop       # Then run the critique/refine loop
    ],
    description="Writes an initial document and then iteratively refines it with critique using an exit tool."
)

// Part of agent.ts --> Follow https://google.github.io/adk-docs/get-started/quickstart/ to learn the setup

import { LoopAgent, LlmAgent, SequentialAgent, FunctionTool } from '@google/adk';
import { z } from 'zod';

// --- Constants ---
const GEMINI_MODEL = "gemini-2.5-flash";
const STATE_INITIAL_TOPIC = "initial_topic";

// --- State Keys ---
const STATE_CURRENT_DOC = "current_document";
const STATE_CRITICISM = "criticism";
// Define the exact phrase the Critic should use to signal completion
const COMPLETION_PHRASE = "No major issues found.";

// --- Tool Definition ---
const exitLoopTool = new FunctionTool({
    name: 'exit_loop',
    description: 'Call this function ONLY when the critique indicates no further changes are needed, signaling the iterative process should end.',
    parameters: z.object({}),
    execute: (input, toolContext) => {
        if (toolContext) {
            console.log(`  [Tool Call] exit_loop triggered by ${toolContext.agentName} with input: ${input}`);
            toolContext.actions.escalate = true;
        }
        return {};
    },
});

// --- Agent Definitions ---

// STEP 1: Initial Writer Agent (Runs ONCE at the beginning)
const initialWriterAgent = new LlmAgent({
    name: "InitialWriterAgent",
    model: GEMINI_MODEL,
    includeContents: 'none',
    // MODIFIED Instruction: Ask for a slightly more developed start
    instruction: `You are a Creative Writing Assistant tasked with starting a story.
    Write the *first draft* of a short story (aim for 2-4 sentences).
    Base the content *only* on the topic provided below. Try to introduce a specific element (like a character, a setting detail, or a starting action) to make it engaging.
    Topic: {{${STATE_INITIAL_TOPIC}}}

    Output *only* the story/document text. Do not add introductions or explanations.
    `,
    description: "Writes the initial document draft based on the topic, aiming for some initial substance.",
    outputKey: STATE_CURRENT_DOC
});

// STEP 2a: Critic Agent (Inside the Refinement Loop)
const criticAgentInLoop = new LlmAgent({
    name: "CriticAgent",
    model: GEMINI_MODEL,
    includeContents: 'none',
    // MODIFIED Instruction: More nuanced completion criteria, look for clear improvement paths.
    instruction: `You are a Constructive Critic AI reviewing a short document draft (typically 2-6 sentences). Your goal is balanced feedback.

    **Document to Review:**

    {{current_document}}


    **Task:**
    Review the document for clarity, engagement, and basic coherence according to the initial topic (if known).

    IF you identify 1-2 *clear and actionable* ways the document could be improved to better capture the topic or enhance reader engagement (e.g., "Needs a stronger opening sentence", "Clarify the character's goal"):
    Provide these specific suggestions concisely. Output *only* the critique text.

    ELSE IF the document is coherent, addresses the topic adequately for its length, and has no glaring errors or obvious omissions:
    Respond *exactly* with the phrase "${COMPLETION_PHRASE}" and nothing else. It doesn't need to be perfect, just functionally complete for this stage. Avoid suggesting purely subjective stylistic preferences if the core is sound.

    Do not add explanations. Output only the critique OR the exact completion.
`,
    description: "Reviews the current draft, providing critique if clear improvements are needed, otherwise signals completion.",
    outputKey: STATE_CRITICISM
});


// STEP 2b: Refiner/Exiter Agent (Inside the Refinement Loop)
const refinerAgentInLoop = new LlmAgent({
    name: "RefinerAgent",
    model: GEMINI_MODEL,
    // Relies solely on state via placeholders
    includeContents: 'none',
    instruction: `You are a Creative Writing Assistant refining a document based on feedback OR exiting the process.
    **Current Document:**

    {{current_document}}

    **Critique/Suggestions:**
    {{criticism}}

    **Task:**
    Analyze the 'Critique/Suggestions'.
    IF the critique is *exactly* "${COMPLETION_PHRASE}":
    You MUST call the 'exit_loop' function. Do not output any text.
    ELSE (the critique contains actionable feedback):
    Carefully apply the suggestions to improve the 'Current Document'. Output *only* the refined document text.

    Do not add explanations. Either output the refined document OR call the exit_loop function.
`,
    tools: [exitLoopTool],
    description: "Refines the document based on critique, or calls exit_loop if critique indicates completion.",
    outputKey: STATE_CURRENT_DOC
});


// STEP 2: Refinement Loop Agent
const refinementLoop = new LoopAgent({
    name: "RefinementLoop",
    // Agent order is crucial: Critique first, then Refine/Exit
    subAgents: [
        criticAgentInLoop,
        refinerAgentInLoop,
    ],
    maxIterations: 5 // Limit loops
});

// STEP 3: Overall Sequential Pipeline
// For ADK tools compatibility, the root agent must be named `root_agent`
export const rootAgent = new SequentialAgent({
    name: "IterativeWritingPipeline",
    subAgents: [
        initialWriterAgent, // Run first to create initial doc
        refinementLoop       // Then run the critique/refine loop
    ],
    description: "Writes an initial document and then iteratively refines it with critique using an exit tool."
});

// ExitLoopArgs defines the (empty) arguments for the ExitLoop tool.
type ExitLoopArgs struct{}

// ExitLoopResults defines the output of the ExitLoop tool.
type ExitLoopResults struct{}

// ExitLoop is a tool that signals the loop to terminate by setting Escalate to true.
func ExitLoop(ctx tool.Context, input ExitLoopArgs) (ExitLoopResults, error) {
    fmt.Printf("[Tool Call] exitLoop triggered by %s \n", ctx.AgentName())
    ctx.Actions().Escalate = true
    return ExitLoopResults{}, nil
}

func main() {
    ctx := context.Background()

    if err := runAgent(ctx, "Write a document about a cat"); err != nil {
        log.Fatalf("Agent execution failed: %v", err)
    }
}

func runAgent(ctx context.Context, prompt string) error {
    model, err := gemini.NewModel(ctx, modelName, &genai.ClientConfig{})
    if err != nil {
        return fmt.Errorf("failed to create model: %v", err)
    }

    // STEP 1: Initial Writer Agent (Runs ONCE at the beginning)
    initialWriterAgent, err := llmagent.New(llmagent.Config{
        Name:        "InitialWriterAgent",
        Model:       model,
        Description: "Writes the initial document draft based on the topic.",
        Instruction: `You are a Creative Writing Assistant tasked with starting a story.
Write the *first draft* of a short story (aim for 2-4 sentences).
Base the content *only* on the topic provided in the user's prompt.
Output *only* the story/document text. Do not add introductions or explanations.`,
        OutputKey: stateDoc,
    })
    if err != nil {
        return fmt.Errorf("failed to create initial writer agent: %v", err)
    }

    // STEP 2a: Critic Agent (Inside the Refinement Loop)
    criticAgentInLoop, err := llmagent.New(llmagent.Config{
        Name:        "CriticAgent",
        Model:       model,
        Description: "Reviews the current draft, providing critique or signaling completion.",
        Instruction: fmt.Sprintf(`You are a Constructive Critic AI reviewing a short document draft.
**Document to Review:**
"""
{%s}
"""
**Task:**
Review the document.
IF you identify 1-2 *clear and actionable* ways it could be improved:
Provide these specific suggestions concisely. Output *only* the critique text.
ELSE IF the document is coherent and addresses the topic adequately:
Respond *exactly* with the phrase "%s" and nothing else.`, stateDoc, donePhrase),
        OutputKey: stateCrit,
    })
    if err != nil {
        return fmt.Errorf("failed to create critic agent: %v", err)
    }

    exitLoopTool, err := functiontool.New(
        functiontool.Config{
            Name:        "exitLoop",
            Description: "Call this function ONLY when the critique indicates no further changes are needed.",
        },
        ExitLoop,
    )
    if err != nil {
        return fmt.Errorf("failed to create exit loop tool: %v", err)
    }

    // STEP 2b: Refiner/Exiter Agent (Inside the Refinement Loop)
    refinerAgentInLoop, err := llmagent.New(llmagent.Config{
        Name:  "RefinerAgent",
        Model: model,
        Instruction: fmt.Sprintf(`You are a Creative Writing Assistant refining a document based on feedback OR exiting the process.
**Current Document:**

"""
{%s}
"""

**Critique/Suggestions:**
{%s}
**Task:**
Analyze the 'Critique/Suggestions'.
IF the critique is *exactly* "%s":
You MUST call the 'exitLoop' function. Do not output any text.
ELSE (the critique contains actionable feedback):
Carefully apply the suggestions to improve the 'Current Document'. Output *only* the refined document text.`, stateDoc, stateCrit, donePhrase),
        Description: "Refines the document based on critique, or calls exitLoop if critique indicates completion.",
        Tools:       []tool.Tool{exitLoopTool},
        OutputKey:   stateDoc,
    })
    if err != nil {
        return fmt.Errorf("failed to create refiner agent: %v", err)
    }

    // STEP 2: Refinement Loop Agent
    refinementLoop, err := loopagent.New(loopagent.Config{
        AgentConfig: agent.Config{
            Name:      "RefinementLoop",
            SubAgents: []agent.Agent{criticAgentInLoop, refinerAgentInLoop},
        },
        MaxIterations: 5,
    })
    if err != nil {
        return fmt.Errorf("failed to create loop agent: %v", err)
    }

    // STEP 3: Overall Sequential Pipeline
    iterativeWriterAgent, err := sequentialagent.New(sequentialagent.Config{
        AgentConfig: agent.Config{
            Name:      appName,
            SubAgents: []agent.Agent{initialWriterAgent, refinementLoop},
        },
    })
    if err != nil {
        return fmt.Errorf("failed to create sequential agent pipeline: %v", err)
    }

import static com.google.adk.agents.LlmAgent.IncludeContents.NONE;

import com.google.adk.agents.LlmAgent;
import com.google.adk.agents.LoopAgent;
import com.google.adk.agents.SequentialAgent;
import com.google.adk.events.Event;
import com.google.adk.runner.InMemoryRunner;
import com.google.adk.sessions.Session;
import com.google.adk.tools.Annotations.Schema;
import com.google.adk.tools.FunctionTool;
import com.google.adk.tools.ToolContext;
import com.google.genai.types.Content;
import com.google.genai.types.Part;
import io.reactivex.rxjava3.core.Flowable;
import java.util.Map;

public class LoopAgentExample {

  // --- Constants ---
  private static final String APP_NAME = "IterativeWritingPipeline";
  private static final String USER_ID = "test_user_456";
  private static final String MODEL_NAME = "gemini-2.0-flash";

  // --- State Keys ---
  private static final String STATE_CURRENT_DOC = "current_document";
  private static final String STATE_CRITICISM = "criticism";

  public static void main(String[] args) {
    LoopAgentExample loopAgentExample = new LoopAgentExample();
    loopAgentExample.runAgent("Write a document about a cat");
  }

  // --- Tool Definition ---
  @Schema(
      description =
          "Call this function ONLY when the critique indicates no further changes are needed,"
              + " signaling the iterative process should end.")
  public static Map<String, Object> exitLoop(@Schema(name = "toolContext") ToolContext toolContext) {
    System.out.printf("[Tool Call] exitLoop triggered by %s \n", toolContext.agentName());
    toolContext.actions().setEscalate(true);
    //  Return empty dict as tools should typically return JSON-serializable output
    return Map.of();
  }

  // --- Agent Definitions ---
  public void runAgent(String prompt) {
    // STEP 1: Initial Writer Agent (Runs ONCE at the beginning)
    LlmAgent initialWriterAgent =
        LlmAgent.builder()
            .model(MODEL_NAME)
            .name("InitialWriterAgent")
            .description(
                "Writes the initial document draft based on the topic, aiming for some initial"
                    + " substance.")
            .instruction(
                """
                    You are a Creative Writing Assistant tasked with starting a story.
                    Write the *first draft* of a short story (aim for 2-4 sentences).
                    Base the content *only* on the topic provided below. Try to introduce a specific element (like a character, a setting detail, or a starting action) to make it engaging.

                    Output *only* the story/document text. Do not add introductions or explanations.
                """)
            .outputKey(STATE_CURRENT_DOC)
            .includeContents(NONE)
            .build();

    // STEP 2a: Critic Agent (Inside the Refinement Loop)
    LlmAgent criticAgentInLoop =
        LlmAgent.builder()
            .model(MODEL_NAME)
            .name("CriticAgent")
            .description(
                "Reviews the current draft, providing critique if clear improvements are needed,"
                    + " otherwise signals completion.")
            .instruction(
                """
                    You are a Constructive Critic AI reviewing a short document draft (typically 2-6 sentences). Your goal is balanced feedback.

                    **Document to Review:**
                    ```
                    {{current_document}}
                    ```

                    **Task:**
                    Review the document for clarity, engagement, and basic coherence according to the initial topic (if known).

                    IF you identify 1-2 *clear and actionable* ways the document could be improved to better capture the topic or enhance reader engagement (e.g., "Needs a stronger opening sentence", "Clarify the character's goal"):
                    Provide these specific suggestions concisely. Output *only* the critique text.

                    ELSE IF the document is coherent, addresses the topic adequately for its length, and has no glaring errors or obvious omissions:
                    Respond *exactly* with the phrase "No major issues found." and nothing else. It doesn't need to be perfect, just functionally complete for this stage. Avoid suggesting purely subjective stylistic preferences if the core is sound.

                    Do not add explanations. Output only the critique OR the exact completion phrase.
                    """)
            .outputKey(STATE_CRITICISM)
            .includeContents(NONE)
            .build();

    // STEP 2b: Refiner/Exiter Agent (Inside the Refinement Loop)
    LlmAgent refinerAgentInLoop =
        LlmAgent.builder()
            .model(MODEL_NAME)
            .name("RefinerAgent")
            .description(
                "Refines the document based on critique, or calls exitLoop if critique indicates"
                    + " completion.")
            .instruction(
                """
                    You are a Creative Writing Assistant refining a document based on feedback OR exiting the process.
                    **Current Document:**
                    ```
                    {{current_document}}
                    ```
                    **Critique/Suggestions:**
                    {{criticism}}

                    **Task:**
                    Analyze the 'Critique/Suggestions'.
                    IF the critique is *exactly* "No major issues found.":
                    You MUST call the 'exitLoop' function. Do not output any text.
                    ELSE (the critique contains actionable feedback):
                    Carefully apply the suggestions to improve the 'Current Document'. Output *only* the refined document text.

                    Do not add explanations. Either output the refined document OR call the exitLoop function.
                """)
            .outputKey(STATE_CURRENT_DOC)
            .includeContents(NONE)
            .tools(FunctionTool.create(LoopAgentExample.class, "exitLoop"))
            .build();

    // STEP 2: Refinement Loop Agent
    LoopAgent refinementLoop =
        LoopAgent.builder()
            .name("RefinementLoop")
            .description("Repeatedly refines the document with critique and then exits.")
            .subAgents(criticAgentInLoop, refinerAgentInLoop)
            .maxIterations(5)
            .build();

    // STEP 3: Overall Sequential Pipeline
    SequentialAgent iterativeWriterAgent =
        SequentialAgent.builder()
            .name(APP_NAME)
            .description(
                "Writes an initial document and then iteratively refines it with critique using an"
                    + " exit tool.")
            .subAgents(initialWriterAgent, refinementLoop)
            .build();

    // Create an InMemoryRunner
    InMemoryRunner runner = new InMemoryRunner(iterativeWriterAgent, APP_NAME);
    // InMemoryRunner automatically creates a session service. Create a session using the service
    Session session = runner.sessionService().createSession(APP_NAME, USER_ID).blockingGet();
    Content userMessage = Content.fromParts(Part.fromText(prompt));

    // Run the agent
    Flowable<Event> eventStream = runner.runAsync(USER_ID, session.id(), userMessage);

    // Stream event response
    eventStream.blockingForEach(
        event -> {
          if (event.finalResponse()) {
            System.out.println(event.stringifyContent());
          }
        });
  }
}