Integrating external agents

external_agents/README.md

@@ -0,0 +1,36 @@
+# Adding external agents to the crew
+
+This example shows how you can add external agents to the crew, on the example of a langchain agent.
+
+You do this by inheriting from the `AgentWrapperParent` class and implementing its abstract methods.
+
+These agents can then be mixed and matched with native CrewAI agents.
+
+An implementation for langchain agent is already provided and can be used as in the following example:
+
+    ```python
+    from langchain import hub
+    from langchain.agents import create_openai_tools_agent
+    from langchain_openai import ChatOpenAI
+    from langchain_community.tools import DuckDuckGoSearchRun
+
+    from crewai.agents.langchain_agent import LangchainCrewAgent
+
+    llm = ChatOpenAI(model="gpt-4-0125-preview", temperature=0)
+    tools = [DuckDuckGoSearchRun()]    
+    researcher_prompt = hub.pull("hwchase17/openai-tools-agent")
+
+    researcher_agent = AgentExecutor(
+        agent=create_openai_tools_agent(llm, tools, researcher_prompt),
+        tools=tools,
+        verbose=True,
+    )
+
+    researcher = LangchainCrewAgent(
+        agent=researcher_agent,
+        tools=[search_tool],
+        role="Senior Research Analyst",
+        allow_delegation=False,
+    )
+
+You need to pass a function that creates the desired agent from the list of its tools, to enable adding other agents as tools at runtime (for delegation).

external_agents/external_agent_demo.py

@@ -0,0 +1,93 @@
+import os
+from typing import Any, List
+import copy
+from uuid import uuid4
+
+from crewai import Agent, Task, Crew
+from crewai.agents.langchain_agent import LangchainCrewAgent
+
+from langchain import hub
+from langchain_community.tools import DuckDuckGoSearchRun
+from langchain_openai import ChatOpenAI
+from langchain.agents import AgentExecutor, create_openai_tools_agent
+
+# Please set the OPENAI_API_KEY environment variable to your OpenAI API key
+# os.environ["OPENAI_API_KEY"] = ...
+
+# You can delete this block if you don't want to use Langsmith
+from langsmith import Client
+
+unique_id = uuid4().hex[0:8]
+os.environ["LANGCHAIN_TRACING_V2"] = "true"
+os.environ["LANGCHAIN_PROJECT"] = f"Tracing Walkthrough - {unique_id}"
+os.environ["LANGCHAIN_ENDPOINT"] = "https://api.smith.langchain.com"
+
+# Replace the LANGCHAIN_API_KEY with your Langsmith API key
+# os.environ["LANGCHAIN_API_KEY"] = ...
+
+client = Client()
+# End of Langsmith block
+
+
+search_tool = DuckDuckGoSearchRun()
+tools = [DuckDuckGoSearchRun()]
+
+researcher_prompt = hub.pull("hwchase17/openai-tools-agent")
+llm = ChatOpenAI(model="gpt-4-0125-preview", temperature=0)
+
+researcher_agent = AgentExecutor(
+    agent=create_openai_tools_agent(llm, tools, researcher_prompt),
+    tools=tools,
+    verbose=True,
+)
+writer_agent = AgentExecutor(
+    agent=create_openai_tools_agent(llm, [], researcher_prompt), tools=[], verbose=True
+)
+
+researcher = LangchainCrewAgent(
+    agent=researcher_agent,
+    tools=[search_tool],
+    role="Senior Research Analyst",
+    allow_delegation=False,
+)
+
+writer = Agent(
+    role="Tech Content Strategist",
+    goal="Craft compelling content on tech advancements",
+    backstory="""You are a renowned Content Strategist, known for your insightful and engaging articles.
+  You transform complex concepts into compelling narratives.""",
+    verbose=True,
+    allow_delegation=True,
+)
+
+# From here onwards it's exactly like the original example
+
+# Create tasks for your agents
+task1 = Task(
+    description="""Conduct a comprehensive analysis of the latest advancements in AI in 2024.
+  Identify key trends, breakthrough technologies, and potential industry impacts.
+  Your final answer MUST be a full analysis report""",
+    agent=researcher,
+)
+
+task2 = Task(
+    description="""Using the insights provided, develop an engaging blog
+  post that highlights the most significant AI advancements.
+  Your post should be informative yet accessible, catering to a tech-savvy audience.
+  Make it sound cool, avoid complex words so it doesn't sound like AI.
+  Your final answer MUST be the full blog post of at least 4 paragraphs.""",
+    agent=writer,
+)
+
+# Instantiate your crew with a sequential process
+crew = Crew(
+    agents=[researcher, writer],
+    tasks=[task2],
+    verbose=2,  # You can set it to 1 or 2 to different logging levels
+)
+
+# Get your crew to work!
+result = crew.kickoff()
+
+print("######################")
+print(result)