Help with Using Yahoo Finance API (get_sec_filing) as a Tool for an Agent in LangFlow

[ealss101]

Hi everyone,

I'm trying to set up a workflow in LangFlow where I use the Yahoo Finance API as a tool for an agent (am using Open AI gpt-40-mini as model). Specifically, I want to use the get_sec_filing method to retrieve SEC filing data.

Here’s my current goal:

1. Use Chat Input as the input for the Yahoo Finance API (e.g., a query like "Get SEC filing for Tesla").
2. Retrieve only the URL from the API response in a format like this: https://cdn.yahoofinance.com/prod/sec-filings/0001318605/000095017023001409/tsla-20221231.htm
3. Pass the retrieved URL into a URL Tool in the Data section of LangFlow for further processing.

End Goal:
I want to chat with my agent and ask it anything about SEC filings. The agent should:

1. Fetch the filing (PDF or .htm) using the retrieved URL.
2. Extract relevant information from the document.
3. Provide me with a precise and cited answer.

Example query to the agent:

"Were there any insider trading incidents with AAPL in their latest SEC 10-k filing? If so, find them and cite them to me!"

The agent should:

• Locate the information in the filing.
• Provide a concise response with citations from the document.

Challenges:
I’m unsure how to:

• Extract just the URL from the yahoo finance API response.
• Pass the URL into the URL Tool for further processing and file extraction.

If there’s a smarter or more efficient way to achieve my end goal (e.g., alternative APIs, tools, or approaches), I’m happy to try that instead. I’m a beginner, but I’m eager to learn and try anything you suggest! Any step-by-step guidance or suggestions would be greatly appreciated. Thank you in advance!

[dosubot[bot]]

Hey @ealss101! I'm here to help you with any bugs, questions, or contributions you have. Let's work on solving this together!

To achieve your goal of using the Yahoo Finance API to retrieve SEC filing data and process it with LangFlow, follow these steps:

1. Set Up the Yahoo Finance Tool:

   • Use the YFinance tool in LangFlow to access the Yahoo Finance API.
   • Configure the tool to use the get_sec_filings method. You can do this by selecting the GET_SEC_FILINGS method from the YahooFinanceMethod enum when setting up the tool ^[1].

2. Extract the URL from the API Response:

   • The API response will contain the SEC filing data, including URLs. Use the ParseJSONDataComponent class in LangFlow to extract specific fields from the JSON response. You can use a JQ query to filter and extract the URL from the response ^[2].

3. Pass the URL to the URL Tool:

   • Once you have extracted the URL, pass it to the URL Tool in LangFlow's Data section. The URL Tool can fetch content from the URL and process it in either "Text" or "Raw HTML" format ^[3].

4. Process the Document:

   • Use the URL Tool to fetch the SEC filing document. If the document is in PDF or HTML format, you can use LangFlow's document processing capabilities to extract relevant information. LangFlow supports various file formats and uses the Unstructured.io API for text extraction ^[4].

5. Configure the Agent:

   • Set up your agent to use the output from the URL Tool. The agent can analyze the extracted text and provide answers to your queries. Ensure the agent is configured to handle the specific types of questions you want to ask about SEC filings.

By following these steps, you can create a workflow in LangFlow that retrieves SEC filing data, processes it, and allows your agent to provide precise and cited answers to your queries.

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Other}

[ealss101]

Issue Description

I am using a JSON parsing tool with jq queries to extract specific fields from my JSON data. However, the output consistently returns None or an empty result, even though the JSON structure appears valid.

JSON Input

Below is the JSON I am using:

[
  {
    "date": "2024-11-01",
    "epochDate": "1730419200",
    "type": "10-K",
    "title": "Periodic Financial Reports",
    "edgarUrl": "https://finance.yahoo.com/sec-filing/AAPL/0000320193-24-000123_320193",
    "exhibits": {
      "EX-10.19": "https://cdn.yahoofinance.com/prod/sec-filings/0000320193/000032019324000123/a10-kexhibit10199282024.htm",
      "10-K": "https://cdn.yahoofinance.com/prod/sec-filings/0000320193/000032019324000123/aapl-20240928.htm"
    },
    "maxAge": "1"
  }
]

Expected Behavior

For a query like .[].edgarUrl, I expect to retrieve the following result:

[
  "https://finance.yahoo.com/sec-filing/AAPL/0000320193-24-000123_320193"
]

Actual Behavior

The query returns None or an empty output.

Steps to Reproduce

1. Input the provided JSON into the parser.
2. Run the jq query .[].edgarUrl (or any similar query).
3. Observe the unexpected result.

Questions

1. Is there anything inherently wrong with the JSON format that would cause the query to fail?
2. Are there specific quirks or limitations with the tool when handling nested structures or field names?
3. Could the use of fields like maxAge (as a string) or deeply nested objects like exhibits cause parsing issues?

Any guidance on how to debug or resolve this issue would be greatly appreciated. Thank you!

---

This description provides all necessary details to help maintainers understand and troubleshoot the issue effectively. Let me know if you'd like to refine it further!