定义结果

“outcome”（结果）告诉会话最终结果应该是什么样子，以及如何衡量其质量。智能体朝着该目标工作，不断进行自我评估和迭代，直到达成结果。

当您定义结果时，“harness”（运行框架）会自动配置一个 grader（评分器），根据 “rubric”（评分标准）评估产出物。评分器使用单独的 “context window”（上下文窗口），以避免受到主智能体实现选择的影响。

评分器会返回一段说明，总结哪些标准通过或未通过，或确认产出物满足评分标准。该反馈会交回给智能体，用于下一次迭代。

创建评分标准

评分标准是一个描述各项标准评分方式的 markdown 文档。评分标准是必需的。

编写有效评分标准的技巧

将评分标准组织为明确、可评分的标准，例如“CSV 包含一个带有数值的 price 列”，而不是“数据看起来不错”。评分器会独立地为每项标准评分，因此模糊的标准会导致评估结果不稳定。

如果您手头没有评分标准，可以尝试向 Claude 提供一个已知优质的产出物示例，让它分析该内容好在哪里，然后将该分析转化为评分标准。这种折中方法通常比从零开始编写标准效果更好。

评分标准示例：

# DCF Model Rubric

## Revenue Projections
- Uses historical revenue data from the last 5 fiscal years
- Projects revenue for at least 5 years forward
- Growth rate assumptions are explicitly stated and reasonable

## Cost Structure
- COGS and operating expenses are modeled separately
- Margins are consistent with historical trends or deviations are justified

## Discount Rate
- WACC is calculated with stated assumptions for cost of equity and cost of debt
- Beta, risk-free rate, and equity risk premium are sourced or justified

## Terminal Value
- Uses either perpetuity growth or exit multiple method (stated which)
- Terminal growth rate does not exceed long-term GDP growth

## Output Quality
- All figures are in a single .xlsx file with clearly labeled sheets
- Key assumptions are on a separate "Assumptions" sheet
- Sensitivity analysis on WACC and terminal growth rate is included

您可以在 user.define_outcome 上以内联文本形式传递评分标准（请参阅创建带有结果的会话），也可以通过 Files API 上传，以便在多个会话中复用。

cURL

curl -fsSL https://api.anthropic.com/v1/files \
  -H "x-api-key: $ANTHROPIC_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "anthropic-beta: managed-agents-2026-04-01" \
  -F file=@/tmp/rubric.md

CLI

ant files upload --file /tmp/rubric.md

Python

import time
from pathlib import Path

from anthropic import Anthropic

client = Anthropic()

RUBRIC = """# DCF Model Rubric

## Revenue Projections
- Uses historical revenue data from the last 5 fiscal years
- Projects revenue for at least 5 years forward

## Output Quality
- All figures are in a single .xlsx file with clearly labeled sheets
"""
Path("/tmp/rubric.md").write_text(RUBRIC)

rubric = client.files.upload(file=Path("/tmp/rubric.md"))
print(f"Uploaded rubric: {rubric.id}")

TypeScript

import { writeFile, readFile } from "node:fs/promises";

import Anthropic from "@anthropic-ai/sdk";
import { toFile } from "@anthropic-ai/sdk";

const client = new Anthropic();

const RUBRIC = `# DCF Model Rubric

## Revenue Projections
- Uses historical revenue data from the last 5 fiscal years
- Projects revenue for at least 5 years forward

## Output Quality
- All figures are in a single .xlsx file with clearly labeled sheets
`;
await writeFile("/tmp/rubric.md", RUBRIC);

const rubric = await client.files.upload({
  file: await toFile(readFile("/tmp/rubric.md"), "/tmp/rubric.md")
});
console.log(`Uploaded rubric: ${rubric.id}`);

C#

using Anthropic;
using Anthropic.Models.Beta.Agents;
using Anthropic.Models.Beta.Environments;
using Anthropic.Models.Beta.Sessions;
using Anthropic.Models.Beta.Sessions.Events;
using Anthropic.Models.Files;

var client = new AnthropicClient();

const string Rubric = """
    # DCF Model Rubric

    ## Revenue Projections
    - Uses historical revenue data from the last 5 fiscal years
    - Projects revenue for at least 5 years forward

    ## Output Quality
    - All figures are in a single .xlsx file with clearly labeled sheets
    """;
await File.WriteAllTextAsync("/tmp/rubric.md", Rubric);

var rubric = await client.Files.Upload(new()
{
    File = File.OpenRead("/tmp/rubric.md"),
});
Console.WriteLine($"Uploaded rubric: {rubric.ID}");

Go

package main

import (
	"context"
	"fmt"
	"io"
	"os"
	"time"

	"github.com/anthropics/anthropic-sdk-go/v2"
)

const rubric = `# DCF Model Rubric

## Revenue Projections
- Uses historical revenue data from the last 5 fiscal years
- Projects revenue for at least 5 years forward

## Output Quality
- All figures are in a single .xlsx file with clearly labeled sheets
`

func main() {
	ctx := context.Background()
	client := anthropic.NewClient()

	if err := os.WriteFile("/tmp/rubric.md", []byte(rubric), 0o644); err != nil {
		panic(err)
	}

	f, err := os.Open("/tmp/rubric.md")
	if err != nil {
		panic(err)
	}

	uploaded, err := client.Files.Upload(ctx, anthropic.FileUploadParams{
		File: anthropic.File(f, "rubric.md", "text/markdown"),
	})
	if err != nil {
		panic(err)
	}
	fmt.Printf("Uploaded rubric: %s\n", uploaded.ID)

Java

import com.anthropic.client.okhttp.AnthropicOkHttpClient;
import com.anthropic.core.http.HttpResponse;
import com.anthropic.models.beta.AnthropicBeta;
import com.anthropic.models.beta.agents.AgentCreateParams;
import com.anthropic.models.beta.agents.BetaManagedAgentsAgentToolset20260401Params;
import com.anthropic.models.beta.agents.BetaManagedAgentsModel;
import com.anthropic.models.beta.environments.BetaCloudConfigParams;
import com.anthropic.models.beta.environments.BetaLimitedNetworkParams;
import com.anthropic.models.beta.environments.EnvironmentCreateParams;
import com.anthropic.models.beta.files.FileListParams;
import com.anthropic.models.beta.sessions.SessionCreateParams;
import com.anthropic.models.beta.sessions.events.BetaManagedAgentsTextRubricParams;
import com.anthropic.models.beta.sessions.events.BetaManagedAgentsUserDefineOutcomeEventParams;
import com.anthropic.models.beta.sessions.events.BetaManagedAgentsUserInterruptEventParams;
import com.anthropic.models.beta.sessions.events.EventSendParams;
import com.anthropic.models.files.FileUploadParams;

import java.io.InputStream;
import java.nio.file.Files;
import java.nio.file.Path;
import java.nio.file.StandardCopyOption;

void main() throws Exception {
    var client = AnthropicOkHttpClient.fromEnv();

    var RUBRIC = """
        # DCF Model Rubric

        ## Revenue Projections
        - Uses historical revenue data from the last 5 fiscal years
        - Projects revenue for at least 5 years forward

        ## Output Quality
        - All figures are in a single .xlsx file with clearly labeled sheets
        """;
    Files.writeString(Path.of("/tmp/rubric.md"), RUBRIC);

    var rubric = client.files().upload(
        FileUploadParams.builder()
            .file(Path.of("/tmp/rubric.md"))
            .build());
    IO.println("Uploaded rubric: " + rubric.id());

PHP

use Anthropic\Client;
use Anthropic\Core\FileParam;

$client = new Client();

$rubricText = <<<'MD'
# DCF Model Rubric

## Revenue Projections
- Uses historical revenue data from the last 5 fiscal years
- Projects revenue for at least 5 years forward

## Output Quality
- All figures are in a single .xlsx file with clearly labeled sheets
MD;
file_put_contents('/tmp/rubric.md', $rubricText);

$rubric = $client->files->upload(
    file: FileParam::fromResource(fopen('/tmp/rubric.md', 'r'), contentType: 'text/markdown'),
);
echo "Uploaded rubric: {$rubric->id}\n";

Ruby

require "anthropic"
require "pathname"

client = Anthropic::Client.new

RUBRIC = <<~MD
  # DCF Model Rubric

  ## Revenue Projections
  - Uses historical revenue data from the last 5 fiscal years
  - Projects revenue for at least 5 years forward

  ## Output Quality
  - All figures are in a single .xlsx file with clearly labeled sheets
MD
File.write("/tmp/rubric.md", RUBRIC)

rubric = client.files.upload(file: Pathname.new("/tmp/rubric.md"))
puts "Uploaded rubric: #{rubric.id}"

创建带有结果的会话

以下示例为现有的智能体和环境（两者均单独创建）创建一个会话，然后发送一个 user.define_outcome 事件。智能体会立即开始工作，无需额外发送用户消息事件。

cURL

# 创建会话
curl -fsSL https://api.anthropic.com/v1/sessions \
  -H "x-api-key: $ANTHROPIC_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "anthropic-beta: managed-agents-2026-04-01" \
  --json '{
    "agent": "'"$AGENT_ID"'",
    "environment_id": "'"$ENVIRONMENT_ID"'",
    "title": "Financial analysis on Costco"
  }'

# 定义结果：SESSION_ID 是创建响应中返回的 ID
# 代理在收到该事件后开始工作
curl -fsSL "https://api.anthropic.com/v1/sessions/$SESSION_ID/events" \
  -H "x-api-key: $ANTHROPIC_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "anthropic-beta: managed-agents-2026-04-01" \
  --json '{
    "events": [
      {
        "type": "user.define_outcome",
        "description": "Build a DCF model for Costco in .xlsx",
        "rubric": {"type": "text", "content": "# DCF Model Rubric\n..."},
        "max_iterations": 5
      }
    ]
  }'
# 或："rubric": {"type": "file", "file_id": "'"$RUBRIC_ID"'"}
# "max_iterations" 为可选项；默认 3，最大 20

CLI

# 创建会话
ant beta:sessions create \
  --agent "$AGENT_ID" \
  --environment-id "$ENVIRONMENT_ID" \
  --title "Financial analysis on Costco"

# 定义结果：SESSION_ID 是创建响应中返回的 ID
# 代理在收到该事件后开始工作
ant beta:sessions:events send --session-id "$SESSION_ID" <<YAML
events:
  - type: user.define_outcome
    description: Build a DCF model for Costco in .xlsx
    rubric: {type: file, file_id: $RUBRIC_ID}
    # or: rubric: {type: text, content: "..."}
    max_iterations: 5  # optional; default 3, max 20
YAML

Python

# 创建会话
session = client.beta.sessions.create(
    agent=agent.id,
    environment_id=environment.id,
    title="Financial analysis on Costco",
)

# 定义结果。代理收到后即开始工作。
client.beta.sessions.events.send(
    session_id=session.id,
    events=[
        {
            "type": "user.define_outcome",
            "description": "Build a DCF model for Costco in .xlsx",
            "rubric": {"type": "text", "content": RUBRIC},
            # 或："rubric": {"type": "file", "file_id": rubric.id},
            "max_iterations": 5,  # optional; default 3, max 20
        }
    ],
)

TypeScript

// 创建会话
const session = await client.beta.sessions.create({
  agent: agent.id,
  environment_id: environment.id,
  title: "Financial analysis on Costco"
});

// 定义结果。智能体收到后即开始工作。
await client.beta.sessions.events.send(session.id, {
  events: [
    {
      type: "user.define_outcome",
      description: "Build a DCF model for Costco in .xlsx",
      rubric: { type: "text", content: RUBRIC },
      // 或：rubric: { type: "file", file_id: rubric.id },
      max_iterations: 5 // optional; default 3, max 20
    }
  ]
});

C#

// 创建会话
var session = await client.Beta.Sessions.Create(new()
{
    Agent = agent.ID,
    EnvironmentID = environment.ID,
    Title = "Financial analysis on Costco",
});

// 定义结果。代理收到后即开始工作。
await client.Beta.Sessions.Events.Send(session.ID, new()
{
    Events =
    [
        new BetaManagedAgentsUserDefineOutcomeEventParams
        {
            Description = "Build a DCF model for Costco in .xlsx",
            Rubric = new BetaManagedAgentsTextRubricParams { Content = Rubric },
            // 或：Rubric = new BetaManagedAgentsFileRubricParams { FileID = rubric.ID },
            MaxIterations = 5, // optional; default 3, max 20
        },
    ],
});

Go

// 创建会话
session, err := client.Beta.Sessions.New(ctx, anthropic.BetaSessionNewParams{
	Agent: anthropic.BetaSessionNewParamsAgentUnion{
		OfString: anthropic.String(agent.ID),
	},
	EnvironmentID: environment.ID,
	Title:         anthropic.String("Financial analysis on Costco"),
})
if err != nil {
	panic(err)
}

// 定义结果。代理收到后即开始工作。
_, err = client.Beta.Sessions.Events.Send(ctx, session.ID, anthropic.BetaSessionEventSendParams{
	Events: []anthropic.BetaManagedAgentsEventParamsUnion{{
		OfUserDefineOutcome: &anthropic.BetaManagedAgentsUserDefineOutcomeEventParams{
			Description: "Build a DCF model for Costco in .xlsx",
			Rubric: anthropic.BetaManagedAgentsUserDefineOutcomeEventParamsRubricUnion{
				OfText: &anthropic.BetaManagedAgentsTextRubricParams{
					Content: rubric,
				},
			},
			// 或：OfFile: &anthropic.BetaManagedAgentsFileRubricParams{FileID: uploaded.ID},
			MaxIterations: anthropic.Int(5), // optional; default 3, max 20
		},
	}},
})
if err != nil {
	panic(err)
}

Java

// 创建会话
var session = client.beta().sessions().create(
    SessionCreateParams.builder()
        .agent(agent.id())
        .environmentId(environment.id())
        .title("Financial analysis on Costco")
        .build());

// 定义结果。代理收到后即开始工作。
client.beta().sessions().events().send(
    session.id(),
    EventSendParams.builder()
        .addEvent(BetaManagedAgentsUserDefineOutcomeEventParams.builder()
            .description("Build a DCF model for Costco in .xlsx")
            .rubric(BetaManagedAgentsTextRubricParams.builder()
                .content(RUBRIC)
                .build())
            // 或：.rubric(BetaManagedAgentsFileRubricParams.builder().fileId(rubric.id()).build())
            .maxIterations(5) // optional; default 3, max 20
            .build())
        .build());

PHP

// 创建会话
$session = $client->beta->sessions->create(
    agent: $agent->id,
    environmentID: $environment->id,
    title: 'Financial analysis on Costco',
);

// 定义结果。代理收到后即开始工作。
$client->beta->sessions->events->send(
    $session->id,
    events: [
        [
            'type' => 'user.define_outcome',
            'description' => 'Build a DCF model for Costco in .xlsx',
            'rubric' => ['type' => 'text', 'content' => $rubricText],
            // 或：'rubric' => ['type' => 'file', 'file_id' => $rubric->id],
            'max_iterations' => 5, // optional; default 3, max 20
        ],
    ],
);

Ruby

# 创建会话
session = client.beta.sessions.create(
  agent: agent.id,
  environment_id: environment.id,
  title: "Financial analysis on Costco"
)

# 定义结果。代理收到后即开始工作。
client.beta.sessions.events.send_(
  session.id,
  events: [
    {
      type: "user.define_outcome",
      description: "Build a DCF model for Costco in .xlsx",
      rubric: {type: "text", content: RUBRIC},
      # 或：rubric: {type: "file", file_id: rubric.id},
      max_iterations: 5 # optional; default 3, max 20
    }
  ]
)

注意

您也可以在创建请求本身中定义结果：在 initial_events (java, php: initialEvents; csharp, go: InitialEvents) 中传入单个 user.define_outcome 事件，即可在一次调用中创建会话并开始朝着结果工作。

结果事件

面向结果的会话的进度会通过事件流呈现。

• agent.* 事件（例如消息和工具使用）显示朝着结果推进的进度。
• span.outcome_evaluation_* 事件仅在面向结果的会话中发出，显示迭代循环的次数以及评分器的反馈过程。
• 您也可以向面向结果的会话发送 user.message 事件，以在智能体工作过程中对其进行引导，但这不是必需的：智能体会自行朝着结果工作，不断迭代，直到成功或用完迭代次数。
• user.interrupt 事件会暂停当前结果的工作，并将 span.outcome_evaluation_end 事件的 result (csharp, go: Result) 标记为 interrupted，使您可以启动新的结果。
• 在最终的结果评估之后，会话可以作为对话式会话继续进行，也可以启动新的结果。会话会保留先前结果的历史记录。

定义结果用户事件

注意

一次仅支持一个结果，但您可以按顺序串联多个结果。为此，请在上一个结果的终止 span.outcome_evaluation_end 事件之后发送新的 user.define_outcome 事件。

这是您为启动结果而发送的事件。收到后会回显该事件，其中包含 processed_at (java, php: processedAt; csharp, go: ProcessedAt) 时间戳和 outcome_id (java: outcomeId; php: outcomeID; csharp, go: OutcomeID)。

{
  "type": "user.define_outcome",
  "description": "Build a DCF model for Costco in .xlsx",
  "rubric": { "type": "file", "file_id": "file_01..." },
  "max_iterations": 5
}

结果评估开始

当评分器针对一个迭代循环开始评估时发出。iteration 字段是一个从 0 开始的修订计数器：0 表示第一次评估，1 表示第一次修订后的重新评估，依此类推。

{
  "type": "span.outcome_evaluation_start",
  "id": "sevt_01def...",
  "outcome_id": "outc_01a...",
  "iteration": 0,
  "processed_at": "2026-03-25T14:01:45Z"
}

结果评估进行中

评分器运行期间发出的心跳事件。评分器的内部推理是不透明的：您只能看到它正在工作，而看不到它在思考什么。

{
  "type": "span.outcome_evaluation_ongoing",
  "id": "sevt_01ghi...",
  "outcome_id": "outc_01a...",
  "iteration": 0,
  "processed_at": "2026-03-25T14:02:10Z"
}

结果评估结束

在一个结果评估周期结束时发出：即评分器完成对一次迭代的评估之后，或在结果处于活动状态时会话被中断时。result 字段指示接下来会发生什么。

结果	下一步
satisfied	会话转换为 idle。
needs_revision	智能体开始新的迭代周期。
max_iterations_reached	在会话转换为 idle 之前，会进行最后一轮确认。不再运行任何评估。
failed	会话转换为 idle。当评分标准不适用于交付物时返回，例如描述与评分标准相互矛盾时。
interrupted	当结果处于活动状态时会话被中断时发出，即使评估尚未开始。如果在中断之前没有触发 outcome_evaluation_start，则 outcome_evaluation_start_id 为空字符串。

{
  "type": "span.outcome_evaluation_end",
  "id": "sevt_01jkl...",
  "outcome_evaluation_start_id": "sevt_01def...",
  "outcome_id": "outc_01a...",
  "result": "satisfied",
  "explanation": "All 12 criteria met: revenue projections use 5 years of historical data, WACC assumptions are stated, sensitivity table is included...",
  "iteration": 0,
  "usage": {
    "input_tokens": 2400,
    "output_tokens": 350,
    "cache_creation_input_tokens": 0,
    "cache_read_input_tokens": 1800
  },
  "processed_at": "2026-03-25T14:03:00Z"
}

检查结果状态

您可以在事件流上监听 span.outcome_evaluation_end，也可以轮询 GET /v1/sessions/{session_id} 并读取 outcome_evaluations[].result。在评估完成之前，result 会报告 pending、running 或 evaluating：

cURL

curl -fsSL "https://api.anthropic.com/v1/sessions/$SESSION_ID" \
  -H "x-api-key: $ANTHROPIC_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "anthropic-beta: managed-agents-2026-04-01"

CLI

ant beta:sessions retrieve --session-id "$SESSION_ID"

Python

session = client.beta.sessions.retrieve(session.id)

for outcome in session.outcome_evaluations:
    print(f"{outcome.outcome_id}: {outcome.result}")
    # outc_01a...: satisfied

TypeScript

const retrieved = await client.beta.sessions.retrieve(session.id);

for (const outcome of retrieved.outcome_evaluations) {
  console.log(`${outcome.outcome_id}: ${outcome.result}`);
  // outc_01a...: satisfied
}

C#

session = await client.Beta.Sessions.Retrieve(session.ID);

foreach (var outcome in session.OutcomeEvaluations)
{
    Console.WriteLine($"{outcome.OutcomeID}: {outcome.Result}");
    // outc_01a...: satisfied
}

Go

session, err = client.Beta.Sessions.Get(ctx, session.ID, anthropic.BetaSessionGetParams{})
if err != nil {
	panic(err)
}

for _, outcome := range session.OutcomeEvaluations {
	fmt.Printf("%s: %s\n", outcome.OutcomeID, outcome.Result)
	// outc_01a...: satisfied
}

Java

var retrieved = client.beta().sessions().retrieve(session.id());

for (var outcome : retrieved.outcomeEvaluations()) {
    IO.println(outcome.outcomeId() + ": " + outcome.result());
    // outc_01a...: satisfied
}

PHP

$session = $client->beta->sessions->retrieve($session->id);

foreach ($session->outcomeEvaluations as $outcome) {
    echo "{$outcome->outcomeID}: {$outcome->result}\n";
    // outc_01a...: satisfied
}

Ruby

session = client.beta.sessions.retrieve(session.id)

session.outcome_evaluations.each do
  puts "#{it.outcome_id}: #{it.result}"
  # outc_01a...: satisfied
end

获取交付物

智能体会将输出文件写入沙箱内的 /mnt/session/outputs/。要获取这些文件，请通过 Files API 列出文件，并将会话 ID 作为 scope_id (java: scopeId; php: scopeID; csharp, go: ScopeID)，然后按 ID 下载。按 scope_id (java: scopeId; php: scopeID; csharp, go: ScopeID) 过滤需要在列表请求中使用 managed-agents-2026-04-01 beta 标头，因此 SDK 和 CLI 示例通过 beta 命名空间发起该调用，并显式传递该标头。文件会在智能体写入完成后不久出现在列表中，有时会在会话进入空闲状态几秒钟之后。如果您预期的文件尚未列出，请稍等片刻后再次列出；一旦它出现在列表中，就表示其上传已完成。

cURL

# 列出此会话生成的文件
# scope_id 过滤需要 managed-agents 测试版
curl -fsSL "https://api.anthropic.com/v1/files?scope_id=$SESSION_ID" \
  -H "x-api-key: $ANTHROPIC_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "anthropic-beta: managed-agents-2026-04-01"

# 按列表中的 ID 下载文件
curl -fsSL "https://api.anthropic.com/v1/files/$FILE_ID/content" \
  -H "x-api-key: $ANTHROPIC_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "anthropic-beta: managed-agents-2026-04-01" \
  -o /tmp/output.txt

CLI

# 列出此会话生成的文件
# scope_id 过滤要求在文件请求中启用 managed-agents beta
ant beta:files list --scope-id "$SESSION_ID" --beta managed-agents-2026-04-01

# 按列表中的 ID 下载文件
ant files download --file-id "$FILE_ID" --output /tmp/output.txt

Python

# 列出此会话生成的文件
# scope_id 过滤要求在文件请求中启用 managed-agents beta
files = client.beta.files.list(scope_id=session.id, betas=["managed-agents-2026-04-01"])
for file in files:
    print(file.id, file.filename)

# 下载文件
if files.data:
    content = client.files.download(files.data[0].id)
    content.write_to_file("/tmp/output.txt")

TypeScript

// 列出此会话生成的文件
// scope_id 过滤要求在文件请求中启用 managed-agents beta
const files = await client.beta.files.list({
  scope_id: session.id,
  betas: ["managed-agents-2026-04-01"]
});
for (const file of files.data) {
  console.log(file.id, file.filename);
}

// 下载文件
if (files.data.length > 0) {
  const content = await client.files.download(files.data[0].id);
  await writeFile("/tmp/output.txt", new Uint8Array(await content.arrayBuffer()));
}

C#

// 列出此会话生成的文件
// （按 scope_id 过滤需要在 files 请求中启用 managed-agents beta）
var files = await client.Beta.Files.List(new()
{
    ScopeID = session.ID,
    Betas = ["managed-agents-2026-04-01"],
});
foreach (var file in files.Items)
{
    Console.WriteLine($"{file.ID} {file.Filename}");
}

// 下载文件
if (files.Items.Count > 0)
{
    using var download = await client.Files.Download(files.Items[0].ID);
    await using var output = File.Create("/tmp/output.txt");
    await (await download.ReadAsStream()).CopyToAsync(output);
}

Go

// 列出此会话生成的文件
// （scope_id 过滤需要在文件请求中启用 managed-agents beta）
files, err := client.Beta.Files.List(ctx, anthropic.BetaFileListParams{
	ScopeID: anthropic.String(session.ID),
	Betas:   []anthropic.AnthropicBeta{anthropic.AnthropicBetaManagedAgents2026_04_01},
})
if err != nil {
	panic(err)
}
for _, file := range files.Data {
	fmt.Println(file.ID, file.Filename)
}

// 下载文件
if len(files.Data) > 0 {
	resp, err := client.Files.Download(ctx, files.Data[0].ID, anthropic.FileDownloadParams{})
	if err != nil {
		panic(err)
	}
	defer resp.Body.Close()
	out, err := os.Create("/tmp/output.txt")
	if err != nil {
		panic(err)
	}
	defer out.Close()
	if _, err := io.Copy(out, resp.Body); err != nil {
		panic(err)
	}
}

Java

// 列出此会话生成的文件
// （scope_id 过滤要求在 files 请求中启用 managed-agents beta）
var files = client.beta().files().list(
    FileListParams.builder()
        .scopeId(session.id())
        .addBeta(AnthropicBeta.MANAGED_AGENTS_2026_04_01)
        .build());
for (var file : files.data()) {
    IO.println(file.id() + " " + file.filename());
}

// 下载文件
if (!files.data().isEmpty()) {
    try (HttpResponse response = client.files().download(files.data().getFirst().id())) {
        try (InputStream body = response.body()) {
            Files.copy(body, Path.of("/tmp/output.txt"), StandardCopyOption.REPLACE_EXISTING);
        }
    }
}

PHP

// 列出此会话生成的文件
// scope_id 过滤要求在文件请求中启用 managed-agents beta
$files = $client->beta->files->list(scopeID: $session->id, betas: ['managed-agents-2026-04-01']);
foreach ($files->getItems() as $file) {
    echo "{$file->id} {$file->filename}\n";
}

// 下载文件
if (count($files->getItems()) > 0) {
    $content = $client->files->download($files->getItems()[0]->id);
    file_put_contents('/tmp/output.txt', $content);
}

Ruby

# 列出此会话生成的文件
# scope_id 过滤要求在文件请求中启用 managed-agents beta
files = client.beta.files.list(scope_id: session.id, betas: ["managed-agents-2026-04-01"])
files.data.each { |file| puts "#{file.id} #{file.filename}" }

# 下载文件
if (first = files.data.first)
  content = client.files.download(first.id)
  File.binwrite("/tmp/output.txt", content.read)
end

后续步骤

• 使用保管库进行身份验证：在创建会话时注册每个用户的凭据。

• 会话事件流：发送事件、流式传输响应，并在执行过程中中断或重定向您的会话。

• 添加文件：上传文件并将其挂载到您的沙箱中，以供读取和处理。

可用性

结果在所有模型上处于 beta 阶段。

API 支持

Managed Agents

• 可用性：可用

状态	云平台
Beta	所有支持的平台

所需的 beta 请求头

Beta 请求头概览

Beta 请求头让您可以在实验性功能和新模型能力成为标准 API 的一部分之前访问它们。

• managed-agents-2026-04-01

数据处理

数据处理概览

项目	零数据保留	HIPAA
状态	不具备资格	不具备资格
详细信息	会话是有状态的资源；对话记录会一直保留，直到您将其删除。适用于所有 Managed Agents 子功能，包括自托管沙箱。	适用于所有 Managed Agents 子功能，包括自托管沙箱。

适用的端点：

• /v1/agents
• /v1/sessions
• /v1/environments