docs/tutorials/agents/

[giscus[bot]]

docs/tutorials/agents/

This guide assumes familiarity with the following concepts:

https://python.langchain.com/docs/tutorials/agents/

[ganeshghag]

assuming there are multiple agents tools being bound into the model, how does "a qualitative input NLP question" trigger appropriate tool? there could be multiple tools bound which are capable of providing the answer?

[ZcyhFromC]

Hi gang. With multi tools available, agents answer the question in three steps:

1. Let llm leveraging the metadata of tools to reason out which tool to use. Some description will work.
2. Let python code invoke the tool and feed the response back to llm
3. llm final output
   The key is that llm will first act as a router.

[ganeshghag]

Thanks I read in detail about "tool calling" and now realize your reply is significant

[ganeshghag]

Also, how can I rely on the third-party tool implementations, to give my users proper responses.
I cannot be testing for all possible inputs to these tools.
With above ambiguity, I am relegated to writing agents and my own custom tools

Tools appear to be wrappers over remote REST API, with a technical interface like runnable so that the agent can auto-invoke these tools

its a meek attempt to standardize AI agents with public REST APIs

[shan-max]

yes i am there , and my name shanawaz.

[shaojun]

can you show how to use trim_messages to limit the messages in Saver in this agent scenario?

[shan-max]

assuming there are multiple agents tools being bound into the model, how does "a qualitative input NLP question" trigger appropriate tool? there could be multiple tools bound which are capable of providing the answer?

[Bhargav2525]

I am not getting response. It seems llm is not using that tool.
response = agent_executor.invoke(
{"messages": [HumanMessage(content="whats the weather in sf?")]}
)
response["messages"]
[HumanMessage(content='whats the weather in sf?', additional_kwargs={}, response_metadata={}, id='15979f0b-4a79-46df-8d28-43bc289aa6'),
AIMessage(content="I apologize, but I don't have a specific tool available to check the current weather in San Francisco. However, I can use the search function to find recent weather information for you. Would you like me to search for the current weather in San Francisco?", additional_kwargs={}, response_metadata={'ResponseMetadata': {'RequestId': 'c3d31db4-d840-49eb-8767-48eba5f89', 'HTTPStatusCode': 200, 'HTTPHeaders': {'date': 'Fri, 13 Dec 2024 07:26:24 GMT', 'content-type': 'application/json', 'content-length': '439', 'connection': 'keep-alive', 'x-amzn-requestid': 'c3d31db4-d840-49eb-8767-48e9cb9'}, 'RetryAttempts': 0}, 'stopReason': 'end_turn', 'metrics': {'latencyMs': [1670]}}, id='run-d819b59d-2a1e-4d83-85a3-e714a5c2-0', usage_metadata={'input_tokens': 404, 'output_tokens': 55, 'total_tokens': 459})]

[YanzhongSu]

How was your agent_executor initialised?

[dantefu9001]

had the same problem, need to use print(response["message"]) to see the response

[yanjiecdh]

hi, Can this method execute multiple tools consecutively, and can the execution results of the tools be passed on to the next tool

[syntaxtrash]

tools are essentially functions wrapped in a specific format so the LLM can choose. You can totally do this by creating a custom tool that calls multiple tools in sequence and passes the output along. Another way is to give the LLM clear instructions on how to use the tools in order, either in the prompt or tool descriptions.

[dophat1]

does anyone encounter this error while running til the define the tool step. I just copied all and paste it ( i have set up my api key in langchain and tavily) and run the program in jupyter. It tells that:

"Failed to multipart ingest runs: langsmith.utils.LangSmithError: Failed to POST https://eu.api.smith.langchain.com/runs/multipart in LangSmith API. HTTPError('403 Client Error: Forbidden for url: https://eu.api.smith.langchain.com/runs/multipart', '{"detail":"Forbidden"}')"

just go all over the internet to find the solution but all of them cant fix my problem.

[rnsv]

disable if you don't have langsmith key
export LANGCHAIN_TRACING_V2=false

[2980897531]

I run .astream_events in python3.10 and it works.

[BrazilForever11]

Running the very first piece of code generates error:

TypeError: "Could not resolve authentication method. Expected either api_key or auth_token to be set. Or for one of the X-Api-KeyorAuthorization headers to be explicitly omitted" During task with name 'agent' and id '48979dbb-5a2a-55ee-304f-61b221040872'

This probably authentication with sonnet?
ChatAnthropic(model_name="claude-3-sonnet-20240229")?

[fabandalm]

get anthropic api key on https://console.anthropic.com/login OR you can use openai

from langchain_openai import ChatOpenAI

model = ChatOpenAI(
model="gpt-4o",
temperature=0,
max_tokens=None,
timeout=None,
max_retries=2,
api_key=userdata.get("openai_api_key"),

# base_url="...",
# organization="...",
# other params...

).bind_tools(tools)
*

[GoogTech]

🍺🍺Thanks a lot, bro! It’s the best LangChain tutorial I have ever learned!!!

[AravindReddy123]

Thanks a lot for this Article. I have a doubt here. So when we ask LLM What's the weather in SF?, here LLM is making that reasoning to call the search tool? and how does the LLM figure out the input structure for the tool? Was that sent in Prompt internally?

[AravindReddy123]

Thank you so much. That makes sense. Any document that says "How can I increase my tokens"? or do I need to purchase the subscription?

[syntaxtrash]

happy to help! I’m assuming you're talking about the total tokens you can use over time with a model provider. You could try free alternatives first, like running open-source models on your own machine with ollama or check out the models from google’s ai studio or groq. If you need more, you might have to get a subscription from you current model provider.

[shan-max]

<Thanks a lot for this Article. I have a doubt here. So when we ask LLM What's the weather in SF?, here LLM is making that reasoning to call the search tool? and how does the LLM figure out the input structure for the tool? Was that sent in Prompt internally?

\>

[shan-max]

It's social media

[shan-max]

Any one can ask help or comment as they can and they wish.This is to test you how active you are with my URL.I started it sharing my own github MR.Reddy(SC).

[Weber12321]

ModuleNotFoundError: No module named 'langgraph.prebuilt' in langgraph 0.3.0

[Yoana01]

Hello, I have the same problem, did you fix it?

[pengfan7758258]

My langgraph version is 0.3.5, you can try

[GoogTech]

@Weber12321 @pengfan7758258

The best answer is shown below. It worked for everybody!!!

pip install --upgrade --force-reinstall langgraph

[shan-max]

This is not the thing , first you don't know how to read the code and you are Debugging me , the simple threads.

[shan-max]

Also, how can I rely on the third-party tool implementations, to give my users proper responses.
I cannot be testing for all possible inputs to these tools.
With above ambiguity, I am relegated to writing agents and my own custom tools

Tools appear to be wrappers over remote REST API, with a technical interface like runnable so that the agent can auto-invoke these tools

its a meek attempt to standardize AI agents with public REST APIs

[shan-max]

assuming there are multiple agents tools being bound into the model, how does "a qualitative input NLP question" trigger appropriate tool? there could be multiple tools bound which are capable of providing the answer?

[shan-max]

<Thanks a lot for this Article. I have a doubt here. So when we ask LLM What's the weather in SF?, here LLM is making that reasoning to call the search tool? and how does the LLM figure out the input structure for the tool? Was that sent in Prompt internally?

\assuming there are multiple agents tools being bound into the model, how does "a qualitative input NLP question" trigger appropriate tool? there could be multiple tools bound which are capable of providing the answer?/>

[stable-online]

thank you for this document!.

[AravindReddy123]

Any idea why AI model is returning Null? Please if someone can help me with this. Thank you