RuleGo
目录

LLM

ai/llm Component: AI text generation component that produces content based on large language model understanding of user inputs, context, or images.

Supports all OpenAI API-compatible large model platforms, such as:

# Configuration

Field Type Description Default
url string API endpoint URL https://ai.gitee.com/v1/
key string API Key
model string Model name (e.g., gpt-3.5-turbo, DeepSeek-R1)
systemPrompt string System prompt defining model behavior and response style. Supports ${} variables
messages []ChatMessage Context/user message list (each with role/user and content)
images []string Images provided to the model for visual understanding
params Params Large model parameters

# Params Structure

Field Type Description Default
temperature float32 Sampling temperature controlling output randomness [0.0, 2.0] 0.0
topP float32 Nucleus sampling: select from top p% tokens [0.0, 1.0] 0.0
presencePenalty float32 Penalize new tokens based on existing presence [0.0, 1.0] 0.0
frequencyPenalty float32 Penalize new tokens based on existing frequency [0.0, 1.0] 0.0
maxTokens int Maximum output length
stop []string Stop sequences
responseFormat string Output format: text/json_object/json_schema text
jsonSchema string JSON Schema definition
keepThink bool Retain reasoning process (text format only) false

# ChatMessage Structure

Field Type Description
role string Message role: user or assistant
content string Content with ${} variable support

# Execution Result

Result replaces msg.Data and flows to next node.The output format is determined by params.responseFormat.

# Configuration Example

{
  "id": "node_2",
  "type": "ai/llm",
  "name": "LLM Request",
  "configuration": {
    "key": "${vars.token}",
    "messages": [
      {"content": "My token: aaabbccc", "role": "user"},
      {"content": "Book 5 tickets for 《Ne Zha 2》", "role": "user"}
    ],
    "model": "Qwen2-7B-Instruct",
    "params": {
      "jsonSchema": "{\"type\":\"object\",\"properties\":{\"name\":{\"type\":\"string\"},\"num\":{\"type\":\"integer\"},\"token\":{\"type\":\"string\"}},\"required\":[\"name\",\"num\",\"token\"]}",
      "responseFormat": "json_schema",
      "temperature": 0.6,
      "topP": 0.75
    },
    "systemPrompt": "You are a ticket assistant. Parse requests into JSON with: token,name,num",
    "url": "https://ai.gitee.com/v1"
  }
}
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

# Application Example

Parse ticket purchase request into JSON

{
  "ruleChain": {
    "id": "bM0H3pgicu_Y",
    "name": "大模型智能体测试",
    "debugMode": false,
    "root": true,
    "disabled": false,
    "configuration": {
      "vars": {
        "token": "xx"
      }
    },
    "additionalInfo": {
      "createTime": "2025/02/13 10:55:48",
      "description": "",
      "layoutX": "280",
      "layoutY": "280",
      "updateTime": "2025/02/13 16:53:30",
      "username": "admin"
    }
  },
  "metadata": {
    "endpoints": [],
    "nodes": [
      {
        "id": "node_2",
        "additionalInfo": {
          "layoutX": 580,
          "layoutY": 270
        },
        "type": "ai/llm",
        "name": "请求大模型，解析票据",
        "debugMode": true,
        "configuration": {
          "key": "${vars.token}",
          "messages": [
            {
              "content": "我的token是:aaabbccc",
              "role": "user"
            },
            {
              "content": "帮我订5张《哪吒2》电影票",
              "role": "user"
            }
          ],
          "model": "Qwen2-7B-Instruct",
          "params": {
            "jsonSchema": "{\n\t\t\t\"type\": \"object\",\n\t\t\t\"properties\": {\n\t\t\t\t\"name\": {\n\t\t\t\t\t\"type\": \"string\"\n\t\t\t\t},\n\t\t\t\t\"num\": {\n\t\t\t\t\t\"type\": \"integer\"\n\t\t\t\t},\n\t\t\t\t\"token\": {\n\t\t\t\t\t\"type\": \"string\"\n\t\t\t\t}\n\t\t\t},\n\t\t\t\"required\": [\"name\", \"num\", \"token\"]\n\t\t}",
            "maxTokens": 0,
            "responseFormat": "json_schema",
            "stop": null,
            "temperature": 0.6,
            "topP": 0.75
          },
          "systemPrompt": "你是一个订票助手，解析用户购票请求，输出Json格式，包含字段：token,name,num",
          "url": "https://ai.gitee.com/v1"
        }
      },
      {
        "id": "node_5",
        "additionalInfo": {
          "layoutX": 840,
          "layoutY": 300
        },
        "type": "jsTransform",
        "name": "请求订票API",
        "debugMode": true,
        "configuration": {
          "jsScript": "return {'msg':msg,'metadata':metadata,'msgType':msgType};"
        }
      }
    ],
    "connections": [
      {
        "fromId": "node_2",
        "toId": "node_5",
        "type": "Success"
      }
    ]
  }
}
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81

Image Recognition

{
  "ruleChain": {
    "id": "hUx-pk6OsMjo",
    "name": "大模型图片内容识别",
    "debugMode": false,
    "root": true,
    "disabled": false,
    "configuration": {
      "vars": {
        "token": "xxx"
      }
    },
    "additionalInfo": {
      "createTime": "2025/02/13 18:50:18",
      "description": "",
      "height": 40,
      "layoutX": "160",
      "layoutY": "250",
      "title": "开始",
      "updateTime": "2025/02/14 21:33:50",
      "username": "admin",
      "width": 240
    }
  },
  "metadata": {
    "endpoints": [],
    "nodes": [
      {
        "id": "node_2",
        "additionalInfo": {
          "background": "",
          "description": "",
          "height": 88,
          "icon": "",
          "layoutX": 460,
          "layoutY": 250,
          "width": 240
        },
        "type": "ai/llm",
        "name": "识别图片内容",
        "debugMode": false,
        "configuration": {
          "description": "",
          "images": [
            "https://rulego.cc/img/architecture_zh.png"
          ],
          "key": "${vars.token}",
          "messages": [
            {
              "content": "解析图片内容",
              "role": "user"
            }
          ],
          "model": "Qwen2-VL-72B",
          "params": {
            "frequencyPenalty": 0,
            "jsonSchema": "",
            "keepThink": false,
            "maxTokens": 0,
            "presencePenalty": 0,
            "responseFormat": "",
            "stop": null,
            "temperature": 0.6,
            "topP": 0.75
          },
          "systemPrompt": "",
          "title": "识别图片内容",
          "url": "https://ai.gitee.com/v1"
        }
      },
      {
        "id": "node_10",
        "type": "log",
        "name": "日志",
        "configuration": {
          "jsScript": "return 'Incoming message:\\n' + JSON.stringify(msg) + '\\nIncoming metadata:\\n' + JSON.stringify(metadata);"
        },
        "debugMode": false,
        "additionalInfo": {
          "layoutX": 700,
          "layoutY": 270
        }
      }
    ],
    "connections": [
      {
        "fromId": "node_2",
        "toId": "node_10",
        "type": "Success"
      }
    ]
  }
}
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
Edit this page on GitHub (opens new window)
Last Updated: 2025/02/17, 13:45:22
WukongimSender
generate-image

← generate-image→

Theme by Vdoing | Copyright © 2023-2025 RuleGo Team | Apache 2.0 License