Gen reply

ReplyGeneratorCogs

ReplyGeneratorCogs(bot: Bot)

Bases: Cog

Parameters:

Name	Type	Description	Default
bot	Bot	The bot instance.	required

Methods:

Name	Description
oai	Generate a reply based on the user's prompt.

Source code in src/cogs/gen_reply.py

def __init__(self, bot: commands.Bot) -> None:
    """Initialize the ReplyGeneratorCogs.

    Args:
        bot (commands.Bot): The bot instance.
    """
    self.bot = bot

bot

bot = bot

oai

oai(
    interaction: Interaction,
    prompt: str = SlashOption(
        description="Enter your prompt.",
        description_localizations={
            Locale.zh_TW: "請輸入提示詞。",
            Locale.ja: "プロンプトを入力してください。",
        },
    ),
    model: str = SlashOption(
        description="Choose a model (default: GPT-4.1).",
        description_localizations={
            Locale.zh_TW: "選擇模型 (預設為 GPT-4.1)",
            Locale.ja: "モデルを選択してください（デフォルトは GPT-4.1）",
        },
        choices=MODEL_CHOICES,
        required=False,
        default="gpt-4.1",
    ),
    stream: bool = SlashOption(
        description="Enable streaming response (default: False).",
        description_localizations={
            Locale.zh_TW: "啟用串流回應 (預設為 False)",
            Locale.ja: "ストリーミング応答を有効にする（デフォルト: False）",
        },
        required=False,
        default=False,
    ),
    image: Optional[Attachment] = SlashOption(
        description="(Optional) Upload an image.",
        description_localizations={
            Locale.zh_TW: "（可選）上傳一張圖片。",
            Locale.ja: "（オプション）画像をアップロードしてください。",
        },
        required=False,
    ),
) -> None

Generate a reply based on the user's prompt.

If the model 'o1' is selected along with an image, an error message is returned since 'o1' does not support image input. The function can either generate a complete response or stream the response in real-time based on the stream parameter.

Parameters:

Name	Type	Description	Default
interaction	Interaction	The interaction object for the command.	required
prompt	str	The prompt text provided by the user.	SlashOption(description='Enter your prompt.', description_localizations={zh_TW: '請輸入提示詞。', ja: 'プロンプトを入力してください。'})
model	str	The selected model, defaults to "gpt-4.1" if not specified.	SlashOption(description='Choose a model (default: GPT-4.1).', description_localizations={zh_TW: '選擇模型 (預設為 GPT-4.1)', ja: 'モデルを選択してください（デフォルトは GPT-4.1）'}, choices=MODEL_CHOICES, required=False, default='gpt-4.1')
stream	bool	Whether to stream the response in real-time, defaults to False.	SlashOption(description='Enable streaming response (default: False).', description_localizations={zh_TW: '啟用串流回應 (預設為 False)', ja: 'ストリーミング応答を有効にする（デフォルト: False）'}, required=False, default=False)
image	Optional[Attachment]	An optional image attachment uploaded by the user.	SlashOption(description='(Optional) Upload an image.', description_localizations={zh_TW: '（可選）上傳一張圖片。', ja: '（オプション）画像をアップロードしてください。'}, required=False)

Source code in src/cogs/gen_reply.py

@nextcord.slash_command(
    name="oai",
    description="Generate a reply based on the given prompt.",
    name_localizations={Locale.zh_TW: "生成文字", Locale.ja: "テキストを生成"},
    description_localizations={
        Locale.zh_TW: "根據提供的提示生成回覆。",
        Locale.ja: "指定されたプロンプトに基づいて応答を生成します。",
    },
    dm_permission=True,
    nsfw=False,
)
async def oai(
    self,
    interaction: Interaction,
    prompt: str = SlashOption(
        description="Enter your prompt.",
        description_localizations={
            Locale.zh_TW: "請輸入提示詞。",
            Locale.ja: "プロンプトを入力してください。",
        },
    ),
    model: str = SlashOption(
        description="Choose a model (default: GPT-4.1).",
        description_localizations={
            Locale.zh_TW: "選擇模型 (預設為 GPT-4.1)",
            Locale.ja: "モデルを選択してください（デフォルトは GPT-4.1）",
        },
        choices=MODEL_CHOICES,
        required=False,
        default="gpt-4.1",
    ),
    stream: bool = SlashOption(
        description="Enable streaming response (default: False).",
        description_localizations={
            Locale.zh_TW: "啟用串流回應 (預設為 False)",
            Locale.ja: "ストリーミング応答を有効にする（デフォルト: False）",
        },
        required=False,
        default=False,
    ),
    image: Optional[nextcord.Attachment] = SlashOption(  # noqa: B008
        description="(Optional) Upload an image.",
        description_localizations={
            Locale.zh_TW: "（可選）上傳一張圖片。",
            Locale.ja: "（オプション）画像をアップロードしてください。",
        },
        required=False,
    ),
) -> None:
    """Generate a reply based on the user's prompt.

    If the model 'o1' is selected along with an image, an error message is returned since 'o1' does not support image input.
    The function can either generate a complete response or stream the response in real-time based on the stream parameter.

    Args:
        interaction (Interaction): The interaction object for the command.
        prompt (str): The prompt text provided by the user.
        model (str): The selected model, defaults to "gpt-4.1" if not specified.
        stream (bool): Whether to stream the response in real-time, defaults to False.
        image (Optional[nextcord.Attachment]): An optional image attachment uploaded by the user.
    """
    await interaction.response.defer()
    attachments = []
    if model not in ["o1", "o1-mini"] and image:
        attachments.append(image.url)

    await interaction.followup.send(content="思考中...")

    try:
        llm_sdk = LLMSDK(model=model)

        if stream:
            # Streaming response
            accumulated_text = f"{interaction.user.mention}\n"
            async for res in llm_sdk.get_oai_reply_stream(
                prompt=prompt, image_urls=attachments
            ):
                if res.choices[0].delta.content:
                    accumulated_text += res.choices[0].delta.content
                    await interaction.edit_original_message(content=accumulated_text)
        else:
            # Non-streaming response
            response = await llm_sdk.get_oai_reply(prompt=prompt, image_urls=attachments)
            final_content = (
                f"{interaction.user.mention}\n{response.choices[0].message.content}"
            )
            await interaction.edit_original_message(content=final_content)

    except Exception:
        error_message = "Error processing the message."
        await interaction.edit_original_message(content=error_message)
        logfire.error("Error in oai", _exc_info=True)