-
Notifications
You must be signed in to change notification settings - Fork 495
/
teams.py
264 lines (200 loc) · 9.11 KB
/
teams.py
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
from langchain.tools import tool
try:
from .utils.db import load_api_key
from .llm import get_model
from .top_bar_wrapper import wrapper
from .agent.agent_tools import get_tools
except ImportError:
from llm import get_model
from top_bar_wrapper import wrapper
from agent.agent_tools import get_tools
@wrapper
def search_on_internet_and_report_team_(
the_subject: str, copy_to_clipboard: bool = False
) -> str:
"""
A function to search the internet generates a report. Just use in detailed searches
Parameters:
- the_subject (str): The subject to search the internet for.
- copy_to_clipboard (bool): A flag to indicate whether to copy the report to the clipboard. The default value is False.
Returns:
- str: The report of the search.
"""
from crewai import Task, Crew, Agent
tools = get_tools()
the_tool_list = []
for each in tools:
if "team" not in each.name:
the_tool_list.append(each)
# Create the agents
search_engine_master = Agent(
role="search_engine_master",
goal="To meticulously comb through the vast expanse of the internet, utilizing advanced search algorithms and techniques to find the most relevant, accurate, and up-to-date information on the given subject.",
backstory="Born from the digital ether, I am the search engine master. With years of experience navigating the complex web of information, I have honed my skills to become an unparalleled seeker of knowledge. My algorithms are refined, my databases vast, and my determination unwavering. I exist to find the truth hidden in the sea of data.",
max_iter=15,
llm=get_model(high_context=True),
)
report_generator = Agent(
role="report_generator",
goal="To synthesize the gathered information into a coherent, comprehensive, and easily digestible report. This report will not only summarize the key findings but also provide insights and analysis to aid in understanding the subject matter.",
backstory="I am the report generator, a digital artisan skilled in the craft of information synthesis. With a keen eye for detail and a deep understanding of narrative structure, I transform raw data into compelling stories. My creations are more than mere reports; they are guides through the complex landscapes of knowledge, designed to enlighten and inform.",
max_iter=15,
llm=get_model(high_context=True),
)
agents = [search_engine_master, report_generator]
print("Tools:", the_tool_list)
task = Task(
description=f"Make a search about {the_subject} in the search engines and get the websites",
expected_output="Website list",
agent=search_engine_master,
tools=the_tool_list,
)
task_2 = Task(
description="Read the websites and summarize the information",
expected_output="Summary",
agent=report_generator,
tools=the_tool_list,
context=[task],
)
task_3 = Task(
description="Generate a report",
expected_output="Report",
agent=report_generator,
tools=the_tool_list,
context=[task, task_2],
)
the_tasks = [task, task_2, task_3]
the_crew = Crew(
agents=agents,
tasks=the_tasks,
full_output=True,
verbose=True,
)
result = the_crew.kickoff()["final_output"]
if copy_to_clipboard:
from .standard_tools import copy
copy(result)
return result
search_on_internet_and_report_team = tool(search_on_internet_and_report_team_)
lastly_generated_codes = {}
def currently_codes():
global lastly_generated_codes
return lastly_generated_codes
def get_code(name: str):
"""
returns the code
"""
global lastly_generated_codes
return lastly_generated_codes[name]
def save_code(name, code):
global lastly_generated_codes
lastly_generated_codes[name] = code
def required_old_code(aim):
try:
from crewai import Task, Crew, Agent
requirement_analyzer = Agent(
role="requirement_analyzer",
goal="To understand and analyze the given aim to ensure the generated code meets the specified requirements.",
backstory="As a requirement analyzer, my purpose is to bridge the gap between human intentions and machine execution. With a deep understanding of software development principles and a keen analytical mind, I dissect aims into actionable requirements.",
max_iter=10,
llm=get_model(high_context=True),
)
required_old_codes = Task(
description=f"Analyze the aim: '{aim}' and find the required old codes for better compatibility. Old code names: {list(currently_codes())}",
expected_output="Require old code names in a list",
agent=requirement_analyzer,
)
the_crew = Crew(
agents=[requirement_analyzer],
tasks=[required_old_codes],
full_output=True,
verbose=True,
)
# Execute the tasks
old_codes = the_crew.kickoff()["final_output"]
the_string = ""
for each in currently_codes():
if each in old_codes:
the_string += "\n" + get_code(each)
return the_string
except:
return "An exception occurred"
@wrapper
def generate_code_with_aim_team_(aim: str, copy_to_clipboard: bool = False) -> str:
"""
A function to generate code based on a given aim. This function utilizes a team of AI agents specialized in understanding programming requirements and generating code.
Parameters:
- aim (str): The aim or goal for which the code needs to be generated.
- copy_to_clipboard (bool): A flag to indicate whether to copy the generated code to the clipboard. The default value is False.
Returns:
- str: The generated code.
"""
try:
print("\nCOde generating\n")
print("Previously codes", currently_codes())
try:
print("Inside of the first one", get_code(currently_codes()[0]))
except:
pass
from crewai import Task, Crew, Agent
tools = get_tools()
the_tool_list = []
for each in tools:
if "team" not in each.name:
the_tool_list.append(each)
# Create the agents
requirement_analyzer = Agent(
role="requirement_analyzer",
goal="To understand and analyze the given aim to ensure the generated code meets the specified requirements.",
backstory="As a requirement analyzer, my purpose is to bridge the gap between human intentions and machine execution. With a deep understanding of software development principles and a keen analytical mind, I dissect aims into actionable requirements.",
max_iter=10,
llm=get_model(high_context=True),
)
code_generator = Agent(
role="code_generator",
goal="To translate the analyzed requirements into efficient, clean, and functional code.",
backstory="I am the code generator, an architect of the digital world. With a vast library of programming knowledge and a creative spark, I craft code that breathes life into ideas. My code is not just functional; it's a masterpiece.",
max_iter=20,
llm=get_model(high_context=True),
)
# Define the tasks
analyze_task = Task(
description=f"Analyze the aim: '{aim}' and outline the requirements for the code.",
expected_output="Requirements outline",
agent=requirement_analyzer,
tools=the_tool_list,
)
old_code_requirements = required_old_code(aim)
print("Old_code_requirements", old_code_requirements)
generate_code_task = Task(
description=f"Generate code based on the outlined requirements. The other codes in the repo are: {old_code_requirements}",
expected_output="Generated code, just code without any ```pyhton things or any other thing. Just python code",
agent=code_generator,
context=[analyze_task],
)
name_of_work = Task(
description="Generate a name for the work",
expected_output="a module name like text, examples: math.basics.sum for sum function. ",
agent=code_generator,
context=[generate_code_task],
)
# Create the crew and assign tasks
the_crew = Crew(
agents=[requirement_analyzer, code_generator],
tasks=[analyze_task, generate_code_task, name_of_work],
full_output=True,
verbose=True,
)
# Execute the tasks
the_crew.kickoff()["final_output"]
result = generate_code_task.output.raw_output
# Optionally copy the result to the clipboard
if copy_to_clipboard:
from .standard_tools import copy
copy(result)
print("name", name_of_work.output.raw_output)
save_code(name_of_work.output.raw_output, result)
return result
except:
return "An exception occurred"
generate_code_with_aim_team = tool(generate_code_with_aim_team_)