程序化工具调用

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
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112import OpenAI from "openai";
import { toResponseInputItems } from "openai/lib/responses/ResponseInputItems";

const client = new OpenAI();

const implementations = {
  get_inventory: async ({ sku }) => ({ sku, available_units: 42 }),
  get_demand: async ({ sku }) => ({ sku, requested_units: 31 }),
};

const tools = [
  {
    type: "function",
    name: "get_inventory",
    description:
      "Return an object with sku (string) and available_units (number).",
    parameters: {
      type: "object",
      properties: { sku: { type: "string" } },
      required: ["sku"],
      additionalProperties: false,
    },
    output_schema: {
      type: "object",
      properties: {
        sku: { type: "string" },
        available_units: { type: "number" },
      },
      required: ["sku", "available_units"],
      additionalProperties: false,
    },
    allowed_callers: ["programmatic"],
    strict: true,
  },
  {
    type: "function",
    name: "get_demand",
    description:
      "Return an object with sku (string) and requested_units (number).",
    parameters: {
      type: "object",
      properties: { sku: { type: "string" } },
      required: ["sku"],
      additionalProperties: false,
    },
    output_schema: {
      type: "object",
      properties: {
        sku: { type: "string" },
        requested_units: { type: "number" },
      },
      required: ["sku", "requested_units"],
      additionalProperties: false,
    },
    allowed_callers: ["programmatic"],
    strict: true,
  },
  { type: "programmatic_tool_calling" },
];

const input = [
  {
    role: "user",
    content: "Compare inventory with demand for sku_123.",
  },
];

while (true) {
  const response = await client.responses.create({
    model: "YOUR_MODEL_ID",
    store: false,
    input,
    tools,
  });

  if (response.status !== "completed") {
    throw new Error(`Response ended with status ${response.status}`);
  }

  // Preserve replayable output, including program and reasoning items.
  input.push(...toResponseInputItems(response.output));

  const calls = response.output.filter((item) => item.type === "function_call");

  if (calls.length === 0) {
    const message = response.output.find((item) => item.type === "message");
    if (message) {
      const refusal = message.content.find((part) => part.type === "refusal");
      console.log(response.output_text || refusal?.refusal || "");
      break;
    }
    continue;
  }

  const outputs = await Promise.all(
    calls.map(async (call) => {
      const run = implementations[call.name];
      if (!run) throw new Error(`Unknown tool: ${call.name}`);

      const result = await run(JSON.parse(call.arguments));
      return {
        type: "function_call_output",
        call_id: call.call_id,
        output: JSON.stringify(result),
        // Preserve caller so the runtime can resume the correct program.
        caller: call.caller,
      };
    })
  );

  input.push(...outputs);
}

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
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179package main

import (
	"context"
	"encoding/json"
	"fmt"

	"github.com/openai/openai-go/v3"
	"github.com/openai/openai-go/v3/responses"
)

type toolArguments struct {
	SKU string `json:"sku"`
}

func main() {
	client := openai.NewClient()
	input := responses.ResponseInputParam{
		responses.ResponseInputItemParamOfMessage(
			"Compare inventory with demand for sku_123.",
			responses.EasyInputMessageRoleUser,
		),
	}
	tools := []responses.ToolUnionParam{
		functionTool(
			"get_inventory",
			"Return an object with sku (string) and available_units (number).",
			"available_units",
		),
		functionTool(
			"get_demand",
			"Return an object with sku (string) and requested_units (number).",
			"requested_units",
		),
		programmaticTool(),
	}

	for {
		response, err := client.Responses.New(context.Background(), responses.ResponseNewParams{
			Model: "gpt-6-astra",
			Store: openai.Bool(false),
			Input: responses.ResponseNewParamsInputUnion{OfInputItemList: input},
			Tools: tools,
		})
		if err != nil {
			panic(err)
		}
		if response.Status != "completed" {
			panic(fmt.Errorf("response ended with status %s", response.Status))
		}

		// Preserve every output item, including program and reasoning items.
		input = append(input, outputAsInput(response.Output)...)

		calls := functionCalls(response.Output)
		if len(calls) == 0 {
			if text, ok := finalMessageText(response); ok {
				fmt.Println(text)
				break
			}
			continue
		}

		for _, call := range calls {
			result, err := runTool(call.Name, call.Arguments)
			if err != nil {
				panic(err)
			}
			output, err := json.Marshal(result)
			if err != nil {
				panic(err)
			}

			toolOutput := responses.ResponseInputItemParamOfFunctionCallOutput(string(output))
			toolOutput.OfFunctionCallOutput.CallID = openai.String(call.CallID)
			caller := call.Caller.AsProgram()
			if caller.CallerID == "" {
				panic("function call is missing its program caller")
			}
			// Preserve caller so the runtime can resume the correct program.
			toolOutput.OfFunctionCallOutput.Caller.OfProgram =
				&responses.ResponseInputItemFunctionCallOutputCallerProgramParam{
					CallerID: caller.CallerID,
				}
			input = append(input, toolOutput)
		}
	}
}

func functionTool(name, description, resultField string) responses.ToolUnionParam {
	parameters := map[string]any{
		"type": "object",
		"properties": map[string]any{
			"sku": map[string]any{"type": "string"},
		},
		"required":             []string{"sku"},
		"additionalProperties": false,
	}
	outputSchema := map[string]any{
		"type": "object",
		"properties": map[string]any{
			"sku":       map[string]any{"type": "string"},
			resultField: map[string]any{"type": "number"},
		},
		"required":             []string{"sku", resultField},
		"additionalProperties": false,
	}
	tool := responses.ToolParamOfFunction(name, parameters, true)
	tool.OfFunction.Description = openai.String(description)
	tool.OfFunction.AllowedCallers = []string{"programmatic"}
	tool.OfFunction.OutputSchema = outputSchema
	return tool
}

func programmaticTool() responses.ToolUnionParam {
	tool := responses.NewToolProgrammaticToolCallingParam()
	return responses.ToolUnionParam{OfProgrammaticToolCalling: &tool}
}

func outputAsInput(
	output []responses.ResponseOutputItemUnion,
) []responses.ResponseInputItemUnionParam {
	input := make([]responses.ResponseInputItemUnionParam, 0, len(output))
	for _, item := range output {
		var converted responses.ResponseInputItemUnion
		if err := json.Unmarshal([]byte(item.RawJSON()), &converted); err != nil {
			panic(err)
		}
		input = append(input, converted.ToParam())
	}
	return input
}

func functionCalls(
	output []responses.ResponseOutputItemUnion,
) []responses.ResponseFunctionToolCall {
	calls := make([]responses.ResponseFunctionToolCall, 0)
	for _, item := range output {
		if item.Type == "function_call" {
			calls = append(calls, item.AsFunctionCall())
		}
	}
	return calls
}

func finalMessageText(response *responses.Response) (string, bool) {
	for _, item := range response.Output {
		if item.Type != "message" {
			continue
		}
		text := response.OutputText()
		if text != "" {
			return text, true
		}
		for _, content := range item.AsMessage().Content {
			if content.Type == "refusal" {
				return content.AsRefusal().Refusal, true
			}
		}
		return "", true
	}
	return "", false
}

func runTool(name, argumentsJSON string) (map[string]any, error) {
	var arguments toolArguments
	if err := json.Unmarshal([]byte(argumentsJSON), &arguments); err != nil {
		return nil, fmt.Errorf("parse %s arguments: %w", name, err)
	}

	switch name {
	case "get_inventory":
		return map[string]any{"sku": arguments.SKU, "available_units": 42}, nil
	case "get_demand":
		return map[string]any{"sku": arguments.SKU, "requested_units": 31}, nil
	default:
		return nil, fmt.Errorf("unknown tool: %s", name)
	}
}

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
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125require "json"
require "openai"

client = OpenAI::Client.new

def get_inventory(sku:)
  {
    sku: sku,
    available_units: 42
  }
end

def get_demand(sku:)
  {
    sku: sku,
    requested_units: 31
  }
end

implementations = {
  "get_inventory" => method(:get_inventory),
  "get_demand" => method(:get_demand)
}
tools = [
  {
    type: :function,
    name: "get_inventory",
    description: "Return an object with sku (string) and available_units (number).",
    parameters: {
      type: :object,
      properties: { sku: { type: :string } },
      required: ["sku"],
      additionalProperties: false
    },
    output_schema: {
      type: :object,
      properties: {
        sku: { type: :string },
        available_units: { type: :number }
      },
      required: %w[sku available_units],
      additionalProperties: false
    },
    allowed_callers: [:programmatic],
    strict: true
  },
  {
    type: :function,
    name: "get_demand",
    description: "Return an object with sku (string) and requested_units (number).",
    parameters: {
      type: :object,
      properties: { sku: { type: :string } },
      required: ["sku"],
      additionalProperties: false
    },
    output_schema: {
      type: :object,
      properties: {
        sku: { type: :string },
        requested_units: { type: :number }
      },
      required: %w[sku requested_units],
      additionalProperties: false
    },
    allowed_callers: [:programmatic],
    strict: true
  },
  { type: :programmatic_tool_calling }
]
input = [
  {
    role: :user,
    content: "Compare inventory with demand for sku_123."
  }
]

loop do
  response = client.responses.create(
    model: "gpt-6-astra",
    store: false,
    input: input,
    tools: tools
  )
  unless response.status == OpenAI::Responses::ResponseStatus::COMPLETED
    raise "Response ended with status #{response.status}"
  end

  # Preserve every output item, including program and reasoning items.
  input.concat(response.output)
  calls = response.output.grep(OpenAI::Models::Responses::ResponseFunctionToolCall)

  if calls.empty?
    message = response.output.find do |item|
      item.is_a?(OpenAI::Models::Responses::ResponseOutputMessage)
    end
    next unless message.is_a?(OpenAI::Models::Responses::ResponseOutputMessage)

    refusal = message.content.find do |content|
      content.is_a?(OpenAI::Models::Responses::ResponseOutputRefusal)
    end
    text = response.output_text
    if text.empty? &&
       refusal.is_a?(OpenAI::Models::Responses::ResponseOutputRefusal)
      text = refusal.refusal
    end
    puts(text)
    break
  end

  calls.each do |call|
    implementation = implementations.fetch(call.name) do
      raise ArgumentError, "Unknown tool: #{call.name}"
    end
    result = implementation.call(**JSON.parse(call.arguments, symbolize_names: true))
    output = {
      type: :function_call_output,
      call_id: call.call_id,
      output: JSON.generate(result)
    }
    # Preserve caller so the runtime can resume the correct program.
    output[:caller] = call.caller_.to_h if call.caller_
    input << output
  end
end