notsobad

[{
    'role': 'system',
    'content': "You are a SQLite expert. Please help to generate a SQL query to answer the question. Your response should ONLY be based on the given context and follow the response guidelines and format instructions. \n===Tables \nCREATE TABLE [Album]\n(\n    [AlbumId] INTEGER  NOT NULL,\n    [Title] NVARCHAR(160)  NOT NULL,\n    [ArtistId] INTEGER  NOT NULL,\n    CONSTRAINT [PK_Album] PRIMARY KEY  ([AlbumId]),\n    FOREIGN KEY ([ArtistId]) REFERENCES [Artist] ([ArtistId]) \n\t\tON DELETE NO ACTION ON UPDATE NO ACTION\n)\n\nCREATE INDEX [IFK_AlbumArtistId] ON [Album] ([ArtistId])\n\nCREATE INDEX [IFK_TrackAlbumId] ON [Track] ([AlbumId])\n\nCREATE TABLE [Track]\n(\n    [TrackId] INTEGER  NOT NULL,\n    [Name] NVARCHAR(200)  NOT NULL,\n    [AlbumId] INTEGER,\n    [MediaTypeId] INTEGER  NOT NULL,\n    [GenreId] INTEGER,\n    [Composer] NVARCHAR(220),\n    [Milliseconds] INTEGER  NOT NULL,\n    [Bytes] INTEGER,\n    [UnitPrice] NUMERIC(10,2)  NOT NULL,\n    CONSTRAINT [PK_Track] PRI

notsobad

# TODO: import the required dependencies
import pandas as pd

# Write code here
sql_query = """
    SELECT country, revenue 
 FROM table_3196ce98cf3ce3223735e1d61022e9d2 

notsobad

Cherry studio use MCP with function call

Client ask question, and convert MCP tool to function call

{
    "model": "deepseek-v3-0324",
    "messages": [{
        "role": "user",
        "content": "当前时间是什么？"
    }],

notsobad

使用function call形式实现MCP调用

这种方式，每个user消息中，带上一个tools，将mcp tools包装成function call提供给LLM。

User msg

{
    "messages": [{
        "content": "You are a helpful AI assistant.
You have access to the following tools:

notsobad

In this environment you have access to a set of tools you can use to answer the user's question. You can use one tool per message, and will receive the result of that tool use in the user's response. You use tools step-by-step to accomplish a given task, with each tool use informed by the result of the previous tool use.

## Tool Use Formatting

Tool use is formatted using XML-style tags. The tool name is enclosed in opening and closing tags, and each parameter is similarly enclosed within its own set of tags. Here's the structure:

<tool_use>
 <name>{tool_name}</name>
 <arguments>{json_arguments}</arguments>

notsobad

// Command click is a chromedp example demonstrating how to use a selector to
// click on an element.
package main

import (
	"context"
	"fmt"
	"log"
	"strings"
	"time"

notsobad

#!/bin/bash

run(){
    cmd=$@
    echo ">> $cmd"
    eval $cmd
    echo
}

run uptime

notsobad

>python3 blackjack.py -h
usage: blackjack.py [-h] -n NROUND [-j BLACKJACK_NUM] [-u PLAYER_HIT_UNTIL] [-s PLAYER_SKIP_LARGE] [--verbose]

Blackjack

optional arguments:
  -h, --help            show this help message and exit
  -n NROUND, --nround NROUND
  -j BLACKJACK_NUM, --blackjack_num BLACKJACK_NUM
  -u PLAYER_HIT_UNTIL, --player_hit_until PLAYER_HIT_UNTIL

notsobad

#!/bin/bash
# Get dns-reverse-proxy from https://github.com/notsobad/dns-reverse-proxy
pushd /xxx/dns-reverse-proxy

LAN_DNS=`systemd-resolve --status | grep 'DNS Servers'|cut -d ':' -f2|xargs`:53
echo "lan dns": $LAN_DNS

sudo ./dns-reverse-proxy\
    -address 0.0.0.0:53 \
    -default 114.114.114.114:53 \

notsobad

#!/bin/bash

function tab () {
    local cmd=""
    local cdto="$PWD"
    local args="$@"

    if [ -d "$1" ]; then
        cdto=`cd "$1"; pwd`
        args="${@:2}"