Repository navigation
Fix empty transcript/scrape results in ChatGPT (v1.3.6) - #34
Merged
Merged
Conversation
ChatGPT gives the model {} for a JSON tool result with a top-level
`content` key, so transcripts and scraped pages never reached it while
metadata did. For ChatGPT only, rename that key to `transcript`
(supadata_transcript, supadata_check_transcript_status) or `markdown`
(supadata_scrape). Other clients keep the REST API's shape.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Deploying with
|
| Status | Name | Latest Commit | Updated (UTC) |
|---|---|---|---|
| ✅ Deployment successful! View logs |
supadata-mcp | d410d7e | Oct 16 2026, 08:33 PM |
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Problem
In ChatGPT,
supadata_transcript,supadata_check_transcript_statusandsupadata_scrapecome back to the model as{}, with no error.supadata_metadataworks. The server sends a correct text result (checked with a direct call to the hosted endpoint using ChatGPT's headers). The one thing the failing results share, and metadata lacks, is a top-levelcontentkey, which ChatGPT appears to read as an MCP tool result's owncontentand drop.Change
For the ChatGPT client only,
renameContentKeyrenames that key before the result is serialized:supadata_transcript,supadata_check_transcript_status:content→transcriptsupadata_scrape:content→markdownEvery other client keeps the same shape as the REST API. Results without
content(job ids, in-progress statuses) and other tools are untouched.Testing
npm test(105 passing), lint,tsc --noEmitsupadata-mcp/1.3.6 (ChatGPT; hosted)🤖 Generated with Claude Code