Change base URL extraction to include scheme - #51
Open
AmarTechWriter wants to merge 1 commit into
Open
Conversation
The base URL passed to LlamaCloudIndex was built using .hostname, which strips the scheme (e.g. "api.cloud.llamaindex.ai" instead of "https://api.cloud.llamaindex.ai"). This produced a malformed URL when the LlamaCloud SDK later concatenated it with a request path, causing a Node ERR_INVALID_URL error on every generateResponse() call. Because this error was silently caught and swallowed, the app fell back to DefaultResponseService's hardcoded template answers instead of surfacing the failure — so response generation appeared to work, but returned generic, topic-mismatched boilerplate instead of real AI-generated, source-grounded answers. Fix: use .origin instead of .hostname to preserve the URL scheme. Also corrected the accompanying comment, which incorrectly described the expected value as a bare hostname. Reproduction: select any populated LlamaCloud index, extract questions from an RFP, and click Generate — responses returned unrelated boilerplate text (e.g. a payload-capacity question answered with content about REST APIs and OAuth) while console.error logged "Falling back to default response due to error" with code: 'ERR_INVALID_URL'. Fixes run-llama#50
Contributor
|
@AmarTechWriter is attempting to deploy a commit to the LlamaIndex Team on Vercel. A member of the Team first needs to authorize it. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The base URL passed to LlamaCloudIndex was built using .hostname, which strips the scheme (e.g. "api.cloud.llamaindex.ai" instead of "https://api.cloud.llamaindex.ai"). This produced a malformed URL when the LlamaCloud SDK later concatenated it with a request path, causing a Node ERR_INVALID_URL error on every generateResponse() call.
Because this error was silently caught and swallowed, the app fell back to DefaultResponseService's hardcoded template answers instead of surfacing the failure — so response generation appeared to work, but returned generic, topic-mismatched boilerplate instead of real AI-generated, source-grounded answers.
Fix: use .origin instead of .hostname to preserve the URL scheme. Also corrected the accompanying comment, which incorrectly described the expected value as a bare hostname.
Reproduction: select any populated LlamaCloud index, extract questions from an RFP, and click Generate — responses returned unrelated boilerplate text (e.g. a payload-capacity question answered with content about REST APIs and OAuth) while console.error logged "Falling back to default response due to error" with code: 'ERR_INVALID_URL'.
Fixes #50