构建计划-执行-反思智能体

POST /_plugins/_ml/connectors/_create
{
    "name": "Amazon Bedrock Claude 3.7-sonnet connector",
    "description": "Connector to Amazon Bedrock service for the Claude model",
    "version": 1,
    "protocol": "aws_sigv4",
    "parameters": {
      "region": "your_aws_region",
      "service_name": "bedrock",
      "model": "us.anthropic.claude-3-7-sonnet-20250219-v1:0"
    },
    "credential": {
      "access_key": "your_aws_access_key",
      "secret_key": "your_aws_secret_key",
      "session_token": "your_aws_session_token"
    },
    "actions": [
      {
        "action_type": "predict",
        "method": "POST",
        "url": "https://bedrock-runtime.${parameters.region}.amazonaws.com/model/${parameters.model}/converse",
        "headers": {
          "content-type": "application/json"
        },
        "request_body": "{ \"system\": [{\"text\": \"${parameters.system_prompt}\"}], \"messages\": [${parameters._chat_history:-}{\"role\":\"user\",\"content\":[{\"text\":\"${parameters.prompt}\"}]}${parameters._interactions:-}]${parameters.tool_configs:-} }"
      }
    ]
}

POST /_plugins/_ml/models/_register
{
    "name": "Bedrock Claude Sonnet model",
    "function_name": "remote",
    "description": "Bedrock Claude 3.7 sonnet model for Plan, Execute and Reflect Agent",
    "connector_id": "your_connector_id"
}

PUT /_plugins/_ml/models/your_model_id
{
  "connector": {
    "client_config": {
      "max_retry_times": -1,
      "retry_backoff_millis": 300,
      "retry_backoff_policy": "exponential_full_jitter"
    }
  }
}

POST _plugins/_ml/agents/_register
{
  "name": "My Plan Execute and Reflect agent with Claude 3.7",
  "type": "plan_execute_and_reflect",
  "description": "this is a test agent",
  "llm": {
    "model_id": "your_llm_model_id_from_step1",
    "parameters": {
      "prompt": "${parameters.question}"
  }},
  "memory": {
    "type": "conversation_index"
  },
  "parameters": {
    "_llm_interface": "bedrock/converse/claude"
  },
  "tools": [
    {
      "type": "ListIndexTool"
    },
    {
      "type": "SearchIndexTool"
    },
    {
      "type": "IndexMappingTool"
    }
  ],
}

tools: [{
    "type": "ListIndexTool",
    "attributes": {
        "input_schema": {
            "type": "object",
            "properties": {
                "indices": {
                    "type": "array",
                    "items": {
                        "type": "string"
                    },
                    "description": "OpenSearch index name list, separated by comma. for example: [\"index1\", \"index2\"], use empty array [] to list all indices in the cluster"
                }
            },
        },
        "strict": false
    }
}]

POST _plugins/_ml/agents/your_agent_id/_execute?async=true
{
  "parameters": {
    "question": "How many flights from Beijing to Seattle?"
  }
}

GET _plugins/_ml/tasks/your_task_id

{
  "task_type": "AGENT_EXECUTION",
  "function_name": "AGENT",
  "state": "COMPLETED",
  "worker_node": [
    "q5yAqa75RM-rv0I67V1VVQ"
  ],
  "create_time": 1746148548710,
  "last_update_time": 1746148706345,
  "is_async": false,
  "response": {
    "memory_id": "bzWQjpYBKhItn1nNYHtu",
    "inference_results": [
      {
        "output": [
          {
            "result": "bzWQjpYBKhItn1nNYHtu",
            "name": "memory_id"
          },
          {
            "result": "cDWQjpYBKhItn1nNYHuS",
            "name": "parent_interaction_id"
          },
          {
            "result": "dTWQjpYBKhItn1nNbHsw",
            "name": "executor_agent_memory_id"
          },
          {
            "result": "YjWQjpYBKhItn1nN6oYk",
            "name": "executor_agent_parent_interaction_id"
          },
          {
            "name": "response",
            "dataAsMap": {
              "response": """# Comprehensive Analysis Report: Flights from Beijing to Seattle

## Executive Summary
After analyzing the OpenSearch sample flight dataset, I found that there are 0 direct flights from Beijing to Seattle in the dataset.

## Analysis Process

### Step 1: Identify Available Data Sources
I began by examining the indices available in the OpenSearch cluster to locate flight-related data. This search revealed one relevant index: `opensearch_dashboards_sample_data_flights`, which contains 13,059 flight records with comprehensive information including origin and destination cities, flight numbers, carriers, and other flight details.

### Step 2: Data Schema Analysis
I analyzed the index structure and confirmed it contains the necessary fields for this investigation, including:
- Origin/destination city names (`OriginCityName`, `DestCityName`) 
- Airport codes (`Origin`, `Dest`)
- Airport IDs (`OriginAirportID`, `DestAirportID`)
- Geographic information for origins and destinations

### Step 3: Query Construction and Execution
I created and executed a search query to find flights where:
- Origin city is Beijing (also checked for "Beijing Capital International Airport" and airport code "PEK")
- Destination city is Seattle (also checked for "Seattle Tacoma International Airport" and airport code "SEA")

### Step 4: Result Verification
To ensure the search was properly constructed, I verified that:
1. Flights from Beijing to other destinations exist in the dataset
2. Flights to Seattle from other origins exist in the dataset

This confirmed that both cities are represented in the data, but no flights connect them directly.

## Key Findings
- Beijing appears as an origin city in the dataset, with flights to destinations including Warsaw, Pittsburgh, Xi'an, Vienna, and Chicago/Rockford
- Seattle appears as both origin and destination in the dataset, with connections to cities like Vienna, Istanbul, New Orleans, St Louis, and Treviso
- The dataset contains 0 flights from Beijing to Seattle

## Conclusion
Based on a comprehensive search of the OpenSearch flight sample dataset, there are 0 flights from Beijing to Seattle in this dataset. While both cities appear in the dataset with connections to other locations, this specific route is not represented in the sample data."""
            }
          }
        ]
      }
    ]
  }
}

GET _plugins/_ml/memory/your_memory_id/messages

GET _plugins/_ml/memory/message/your_message_id/traces

POST _plugins/_ml/agents/your_agent_id/_execute?async=true
{
  "parameters": {
    "question": "your_question",
    "memory_id": "your_memory_id",
  }
}