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
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
| class DifyAgentCallbackHandler:
"""
Dify Agent回调处理器
处理Agent执行过程中的各种回调事件
"""
def __init__(self):
"""初始化回调处理器"""
self.callbacks: List[AgentCallback] = []
self.event_handlers: Dict[str, List[Callable]] = {}
def register_callback(self, callback: 'AgentCallback'):
"""
注册回调处理器
Args:
callback: 回调处理器实例
"""
self.callbacks.append(callback)
def register_event_handler(self, event_type: str, handler: Callable):
"""
注册事件处理器
Args:
event_type: 事件类型
handler: 处理器函数
"""
if event_type not in self.event_handlers:
self.event_handlers[event_type] = []
self.event_handlers[event_type].append(handler)
def on_agent_start(self, agent_config: AgentEntity, query: str):
"""
Agent开始执行回调
Args:
agent_config: Agent配置
query: 用户查询
"""
for callback in self.callbacks:
try:
callback.on_agent_start(agent_config, query)
except Exception as e:
logger.exception(f"回调处理器执行失败 - on_agent_start: {e}")
def on_llm_start(self, prompt_messages: List[PromptMessage], tools: List[PromptMessageTool]):
"""
LLM开始调用回调
Args:
prompt_messages: 提示消息列表
tools: 可用工具列表
"""
for callback in self.callbacks:
try:
callback.on_llm_start(prompt_messages, tools)
except Exception as e:
logger.exception(f"回调处理器执行失败 - on_llm_start: {e}")
def on_llm_chunk(self, chunk: LLMResultChunk):
"""
LLM流式输出块回调
Args:
chunk: LLM结果块
"""
for callback in self.callbacks:
try:
callback.on_llm_chunk(chunk)
except Exception as e:
logger.exception(f"回调处理器执行失败 - on_llm_chunk: {e}")
def on_llm_end(self, llm_result: LLMResult):
"""
LLM调用结束回调
Args:
llm_result: LLM调用结果
"""
for callback in self.callbacks:
try:
callback.on_llm_end(llm_result)
except Exception as e:
logger.exception(f"回调处理器执行失败 - on_llm_end: {e}")
def on_tool_start(self, tool_name: str, tool_input: Dict[str, Any]):
"""
工具开始调用回调
Args:
tool_name: 工具名称
tool_input: 工具输入参数
"""
for callback in self.callbacks:
try:
callback.on_tool_start(tool_name, tool_input)
except Exception as e:
logger.exception(f"回调处理器执行失败 - on_tool_start: {e}")
# 触发事件处理器
self._trigger_event_handlers("tool_start", {
"tool_name": tool_name,
"tool_input": tool_input
})
def on_tool_end(self, tool_name: str, tool_result: ToolInvokeMessage):
"""
工具调用结束回调
Args:
tool_name: 工具名称
tool_result: 工具调用结果
"""
for callback in self.callbacks:
try:
callback.on_tool_end(tool_name, tool_result)
except Exception as e:
logger.exception(f"回调处理器执行失败 - on_tool_end: {e}")
# 触发事件处理器
self._trigger_event_handlers("tool_end", {
"tool_name": tool_name,
"tool_result": tool_result.dict() if hasattr(tool_result, 'dict') else str(tool_result)
})
def on_tool_error(self, tool_name: str, error: Exception):
"""
工具调用错误回调
Args:
tool_name: 工具名称
error: 错误信息
"""
for callback in self.callbacks:
try:
callback.on_tool_error(tool_name, error)
except Exception as e:
logger.exception(f"回调处理器执行失败 - on_tool_error: {e}")
# 触发事件处理器
self._trigger_event_handlers("tool_error", {
"tool_name": tool_name,
"error": str(error)
})
def on_agent_finish(self, final_answer: str, reasoning_history: List[AgentReasoningStep]):
"""
Agent执行完成回调
Args:
final_answer: 最终答案
reasoning_history: 推理历史
"""
for callback in self.callbacks:
try:
callback.on_agent_finish(final_answer, reasoning_history)
except Exception as e:
logger.exception(f"回调处理器执行失败 - on_agent_finish: {e}")
def on_agent_error(self, error: Exception, reasoning_history: List[AgentReasoningStep]):
"""
Agent执行错误回调
Args:
error: 错误信息
reasoning_history: 推理历史
"""
for callback in self.callbacks:
try:
callback.on_agent_error(error, reasoning_history)
except Exception as e:
logger.exception(f"回调处理器执行失败 - on_agent_error: {e}")
def _trigger_event_handlers(self, event_type: str, event_data: Dict[str, Any]):
"""
触发事件处理器
Args:
event_type: 事件类型
event_data: 事件数据
"""
handlers = self.event_handlers.get(event_type, [])
for handler in handlers:
try:
handler(event_data)
except Exception as e:
logger.exception(f"事件处理器执行失败 - {event_type}: {e}")
class AgentCallback(ABC):
"""Agent回调抽象基类"""
@abstractmethod
def on_agent_start(self, agent_config: AgentEntity, query: str):
"""Agent开始执行"""
pass
@abstractmethod
def on_llm_start(self, prompt_messages: List[PromptMessage], tools: List[PromptMessageTool]):
"""LLM开始调用"""
pass
@abstractmethod
def on_llm_chunk(self, chunk: LLMResultChunk):
"""LLM流式输出块"""
pass
@abstractmethod
def on_llm_end(self, llm_result: LLMResult):
"""LLM调用结束"""
pass
@abstractmethod
def on_tool_start(self, tool_name: str, tool_input: Dict[str, Any]):
"""工具开始调用"""
pass
@abstractmethod
def on_tool_end(self, tool_name: str, tool_result: ToolInvokeMessage):
"""工具调用结束"""
pass
@abstractmethod
def on_tool_error(self, tool_name: str, error: Exception):
"""工具调用错误"""
pass
@abstractmethod
def on_agent_finish(self, final_answer: str, reasoning_history: List[AgentReasoningStep]):
"""Agent执行完成"""
pass
@abstractmethod
def on_agent_error(self, error: Exception, reasoning_history: List[AgentReasoningStep]):
"""Agent执行错误"""
pass
class LoggingAgentCallback(AgentCallback):
"""日志记录回调处理器"""
def __init__(self, logger_name: str = "agent"):
self.logger = logging.getLogger(logger_name)
def on_agent_start(self, agent_config: AgentEntity, query: str):
self.logger.info(f"Agent开始执行 - 策略: {agent_config.strategy}, 查询: {query}")
def on_llm_start(self, prompt_messages: List[PromptMessage], tools: List[PromptMessageTool]):
tool_names = [tool.name for tool in tools]
self.logger.info(f"LLM调用开始 - 工具: {tool_names}")
def on_llm_chunk(self, chunk: LLMResultChunk):
# 通常不记录每个chunk，避免日志过多
pass
def on_llm_end(self, llm_result: LLMResult):
usage = llm_result.usage
self.logger.info(f"LLM调用完成 - 令牌: {usage.total_tokens}, 成本: ${usage.total_price:.4f}")
def on_tool_start(self, tool_name: str, tool_input: Dict[str, Any]):
self.logger.info(f"工具调用开始 - {tool_name}: {tool_input}")
def on_tool_end(self, tool_name: str, tool_result: ToolInvokeMessage):
result_preview = str(tool_result.message)[:100] + "..." if len(str(tool_result.message)) > 100 else str(tool_result.message)
self.logger.info(f"工具调用完成 - {tool_name}: {result_preview}")
def on_tool_error(self, tool_name: str, error: Exception):
self.logger.error(f"工具调用失败 - {tool_name}: {error}")
def on_agent_finish(self, final_answer: str, reasoning_history: List[AgentReasoningStep]):
self.logger.info(f"Agent执行完成 - 迭代次数: {len(reasoning_history)}")
def on_agent_error(self, error: Exception, reasoning_history: List[AgentReasoningStep]):
self.logger.error(f"Agent执行失败 - 迭代次数: {len(reasoning_history)}, 错误: {error}")
class MetricsAgentCallback(AgentCallback):
"""性能指标回调处理器"""
def __init__(self, metrics_collector: 'AgentMetricsCollector'):
self.metrics_collector = metrics_collector
def on_agent_start(self, agent_config: AgentEntity, query: str):
self.metrics_collector.record_agent_start(agent_config.strategy)
def on_llm_start(self, prompt_messages: List[PromptMessage], tools: List[PromptMessageTool]):
self.metrics_collector.record_llm_call()
def on_llm_chunk(self, chunk: LLMResultChunk):
pass # 流式输出不单独记录
def on_llm_end(self, llm_result: LLMResult):
self.metrics_collector.record_token_usage(llm_result.usage)
def on_tool_start(self, tool_name: str, tool_input: Dict[str, Any]):
self.metrics_collector.record_tool_call(tool_name)
def on_tool_end(self, tool_name: str, tool_result: ToolInvokeMessage):
self.metrics_collector.record_tool_success(tool_name)
def on_tool_error(self, tool_name: str, error: Exception):
self.metrics_collector.record_tool_error(tool_name, error)
def on_agent_finish(self, final_answer: str, reasoning_history: List[AgentReasoningStep]):
self.metrics_collector.record_agent_success(len(reasoning_history))
def on_agent_error(self, error: Exception, reasoning_history: List[AgentReasoningStep]):
self.metrics_collector.record_agent_error(error, len(reasoning_history))
|