with one click
baidu-search
百度AI搜索 - 基于千帆平台的智能搜索服务 提供AI增强的搜索结果,包含引用来源标注
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Menu
百度AI搜索 - 基于千帆平台的智能搜索服务 提供AI增强的搜索结果,包含引用来源标注
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
| name | baidu-search |
| version | 1.0.0 |
| status | implemented |
| description | 百度AI搜索 - 基于千帆平台的智能搜索服务 提供AI增强的搜索结果,包含引用来源标注 |
| tags | ["search","baidu","qianfan","web-search"] |
| requires | {"tools":["Bash"]} |
| env | ["BAIDU_QIANFAN_API_KEY"] |
百度AI搜索是基于百度千帆平台的智能搜索服务,提供AI增强的搜索结果,支持:
Endpoint: https://qianfan.baidubce.com/v2/ai_search/chat/completions
Authentication: Bearer Token
import os
import requests
api_key = os.environ.get("BAIDU_QIANFAN_API_KEY")
headers = {
"Content-Type": "application/json",
"Authorization": f"Bearer {api_key}"
}
def baidu_search(query: str) -> dict:
"""
Perform Baidu AI Search.
Args:
query: Search query string
Returns:
dict with keys: result, references, is_safe
"""
import os
import requests
# 加载 .env 文件
from dotenv import load_dotenv
load_dotenv()
api_key = os.environ.get("BAIDU_QIANFAN_API_KEY")
if not api_key:
raise ValueError("BAIDU_QIANFAN_API_KEY not set")
url = "https://qianfan.baidubce.com/v2/ai_search/chat/completions"
payload = {
"messages": [{"role": "user", "content": query}],
"model": "ernie-4.5-turbo-32k",
"search_source": "baidu_search_v2",
"enable_corner_markers": True,
"stream": False
}
response = requests.post(url, headers=headers, json=payload, timeout=120)
response.raise_for_status()
return response.json()
| Parameter | Type | Default | Description |
|---|---|---|---|
query | string | required | Search query |
model | string | ernie-4.5-turbo-32k | Model to use |
search_source | string | baidu_search_v2 | Search engine version |
enable_deep_search | bool | False | Enable deep search mode |
enable_corner_markers | bool | True | Enable citation markers |
enable_followup_queries | bool | False | Enable follow-up suggestions |
search_recency_filter | string | None | Time filter: week/month/semiyear/year |
temperature | float | 0.11 | Sampling temperature |
top_p | float | 0.55 | Sampling top_p |
{
"choices": [{
"finish_reason": "stop",
"message": {
"content": "AI生成的搜索答案...",
"role": "assistant"
}
}],
"is_safe": true,
"references": [
{
"id": 1,
"title": "来源标题",
"url": "https://example.com",
"content": "来源内容摘要",
"website": "网站名称",
"date": "2026-03-22"
}
]
}
result = baidu_search("今天北京天气怎么样")
print(result["choices"][0]["message"]["content"])
def baidu_deep_search(query: str) -> dict:
payload = {
"messages": [{"role": "user", "content": query}],
"model": "ernie-4.5-turbo-32k",
"enable_deep_search": True,
"enable_corner_markers": True
}
# ... make request
def baidu_search_recent(query: str, days: int = 7) -> dict:
filter_map = {7: "week", 30: "month", 180: "semiyear", 365: "year"}
recency = filter_map.get(days, "month")
payload = {
"messages": [{"role": "user", "content": query}],
"model": "ernie-4.5-turbo-32k",
"search_recency_filter": recency
}
# ... make request
| Variable | Required | Description |
|---|---|---|
BAIDU_QIANFAN_API_KEY | Yes | Baidu Qianfan API Key |
try:
result = baidu_search("search query")
except requests.exceptions.Timeout:
print("Request timed out")
except requests.exceptions.HTTPError as e:
print(f"HTTP error: {e.response.status_code}")
except ValueError as e:
print(f"Configuration error: {e}")
bce-v3/ALTAK-xxx/xxx智能体工作流:单入口设计,所有任务从 router 开始 TRIGGER when: 开发、修复、规划、分析、审查、调研、实施 DO NOT TRIGGER when: 简单闲聊
完成阶段 - 收尾工作、自反思和状态更新 当前版本对齐 .workflow_state.json 状态管理 v1.2: 新增可选 learn 技能触发建议与总结摘要复用
调试阶段 - 系统化问题定位和修复 5步调试法:闻味道 → 揪头发 → 照镜子 → 执行 → 复盘 包含压力升级机制和7项检查清单 v1.2: 强化 context-aware 激活、memory_hints 复用与失败升级事件记录
执行阶段 - TDD 开发循环和代码实现 当前版本对齐 .specs/<feature>/tasks.md、.contract.json、run_tracker.py 和 step_recorder.py Reflexion 强化: mid-task reflection checkpoint + AgentSys 子 agent schema 验证 v1.3: 新增 memory_hints/memory_query/memory_intent 复用与阶段摘要回流
深度探索阶段 - 苏格拉底式追问挖掘用户深层想法 通过迭代追问发现谬误、局限、潜能和潜意识中的构思
产品咨询阶段 - 在开始编码前重构产品想法 问正确的问题,确保理解用户真正想要什么