Error reported when using APIChain.from_llm.and-api_docs is not in the allowed domains

[lllyyyccc]

Checked other resources

• I added a very descriptive title to this question.
• I searched the LangChain documentation with the integrated search.
• I used the GitHub search to find a similar question and didn't find it.

Commit to Help

• I commit to help with one of those options 👆

Example Code

llm = ChatOpenAI(model='gpt-4-turbo', temperature=0)
        chain = APIChain.from_llm_and_api_docs(
            llm,
            salesDocs.TREND_DATA_DOCS,
            headers=headers,
            # verbose=True,
            api_url_prompt=promptUrl.API_URL_PROMPT,
            api_response_prompt=promptUrl.API_RESPONSE_PROMPT,
            limit_to_domains=[api_domains],
        )
        try:
            res = chain.invoke(self.userInput)
            return res['output']
        except Exception as e:
            return e

Description

Error reported when using APIChain.from_llm.and-api_docs is not in the allowed domains。
Why is the inference process requesting an interface together instead of a separate interface address requesting an interface ？
build API URL:

• from: llm
• start_time: 20230521
• end_time: 20230726
• field: sales_amount
• date_type: 3
• get_type: 1

The complete API URL will be as follows：

http://www.example.com/example.example.example/example?example=example&example=20230521&example=20230726&example=example&example=3&example=1

This URL will be used to obtain comprehensive trend analysis of sales within the specified date range, including current period data, month on month, and year-on-year data。 is not in the allowed domains: ['http://www.example.com/']

System Info

System Information

OS: Windows
OS Version: 10.0.22631
Python Version: 3.11.5 (tags/v3.11.5:cce6ba9, Aug 24 2023, 14:38:34) [MSC v.1936 64 bit (AMD64)]

Package Information

langchain_core: 0.1.43
langchain: 0.1.16
langchain_community: 0.0.33
langsmith: 0.1.48
langchain_openai: 0.1.3
langchain_text_splitters: 0.0.1
langchainhub: 0.1.14

Packages not installed (Not Necessarily a Problem)

The following packages were not found:

langgraph
langserve

[dosubot[bot]]

The issue you're encountering with APIChain.from_llm_and_api_docs not recognizing the domain in limit_to_domains likely stems from how the api_domains variable is being passed. Ensure that api_domains is a flat list of domain strings, not a nested list or a single string. For example, it should look like this:

api_domains = ["http://www.example.com"]

Then, pass api_domains directly to limit_to_domains without wrapping it in another list:

chain = APIChain.from_llm_and_api_docs(
    llm,
    salesDocs.TREND_DATA_DOCS,
    headers=headers,
    api_url_prompt=promptUrl.API_URL_PROMPT,
    api_response_prompt=promptUrl.API_RESPONSE_PROMPT,
    limit_to_domains=api_domains,
)

This adjustment ensures that the domain check is performed correctly against the list of allowed domains.

Regarding your question about the interface address, the APIChain.from_llm_and_api_docs method is designed for use with a single API endpoint. If your scenario requires interacting with multiple interfaces, consider creating separate instances of APIChain for each endpoint to manage these requests effectively.

Sources

• libs/langchain/langchain/chains/api/base.py

About Dosu

This response is meant to be useful and save you time. It is not meant to be a precise solution, but rather a starting point for your own research.

Help me be more useful! Please leave a 👍 if this is helpful and 👎 if it is irrelevant.

If you want to continue the conversation, start your reply with @dosu-bot.

[lllyyyccc]

It's not a problem with limit_to_domains, it's the inference process that comes with the requested interface

[edubezerra]

It took me about ~4 hours to find a solution to this problem. The solution was to explicitly define in the template that only the URL of the end point should be returned and no extra text explaining why the URL was built the way it was.