1

pip install openai

1

2

OPENAI_API_KEY = sk-xxxxx

OPENAI_BASE_URL = https://api.openai.com/v1

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

import os

from flask import Flask, jsonify

from openai import OpenAI

config = configparser.ConfigParser()

config.read("config.cfg", encoding="utf-8")

OPENAI_API_KEY = config.get("default", "OPENAI_API_KEY", fallback=None)

OPENAI_BASE_URL = config.get("default", "OPENAI_BASE_URL", fallback=None)

@app.route("/gpt_test")

def gpt_test():

    """

    简单调用一次 GPT，返回一个固定问题的回答

    """

    if not OPENAI_API_KEY:

        return jsonify({"error": "OPENAI_API_KEY 未配置"}), 500

    try:

        # 这里用的是 chat.completions.create 风格

        resp = client.chat.completions.create(

            model="gpt-4.1-mini",  # 或者你有的任意模型，比如 gpt-4.1, gpt-4o 等

            messages=[

                {"role": "system", "content": "你是一个简洁回答的助手。"},

                {"role": "user", "content": "简单用一句话介绍一下你自己。"},

            ],

        )

        answer = resp.choices[0].message.content

        return jsonify({"answer": answer})

    except Exception as e:

        print("GPT 调用异常:", repr(e))

        return jsonify({"error": str(e)}), 500

1

pip install dashscope

1

2

3

4

5

6

7

8

9

10

11

12

13

14

ALIYUN_API_KEY = config.get("default", "ALIYUN_API_KEY", fallback=None)

    @app.route("/llm_test/")

    def llm_test():

        """测试与大模型的对话功能"""

        try:

            messages = [

                {'role': 'system', 'content': 'You are a helpful assistant.'},

                {'role': 'user', 'content': '你是谁？'}

            ]

            answer = chat_with_model(messages)

            return jsonify({"answer": answer})

        except Exception as e:

            print("LLM error:", repr(e))

            return jsonify({"error": str(e)}), 500

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

from pathlib import Path

# run.py 所在目录

BASE_DIR = Path(__file__).resolve().parent

PROMPT_DIR = BASE_DIR / "app" / "prompt_store"

PROMPT_LATEST_FILE = PROMPT_DIR / "doc-llm-latest.md"

def load_latest_prompt() -> str | None:

    """读取 doc-llm-latest.md 的内容"""

    try:

        with PROMPT_LATEST_FILE.open("r", encoding="utf-8") as f:

            return f.read()

    except FileNotFoundError:

        print(f"[WARN] Prompt file not found: {PROMPT_LATEST_FILE}")

        return None

    except Exception as e:

        print(f"[ERROR] Failed to read prompt: {e!r}")

        return None

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

@app.route("/llm_with_prompt/")

def llm_with_prompt():

    """使用最新的Prompt与大模型对话"""

    prompt = load_latest_prompt()

    if not prompt:

        return jsonify({"error": "No prompt available"}), 500

    try:

        messages = [

            {

                'role': 'system',

                'content': prompt

            },

            {

                'role': 'user',

                'content': "请用一两句话，概括一下这个文档测试规范的核心目标。"

            }

        ]

        answer = chat_with_model(messages)

        return jsonify({"answer": answer})

    except Exception as e:

        print("LLM with prompt error:", repr(e))

        return jsonify({"error": str(e)}), 500