Refer to the following troubleshooting procedure when you encounter a conversion exception:
- Checking the view configuration: Confirm that the currently active view prompt meets expectations, and switch back to raw view to see the quality of the original HTML if necessary.
- Model Selection: change to a more powerful model in the configuration file (e.g. upgrade gpt-4.1-nano to gpt-4)
- Cue word optimization::
- Add specific instruction: "Keep all code examples"
- Restricted output format: "Present key points in a secondary list"
- network diagnosis: Verification of network latency and response status due to reliance on external APIs
- Content chunking: Large web pages may try to handle key section URLs separately first
Advanced debugging suggests checking out the issues section of the GitHub repository, FAQs include:
- Misaligned table data conversion (requires specific markdown prompt words)
- Non-English content processing (language parameters to be configured)
- Paging content splicing (requires continuous URL auto-crawl function)
The developer community is usually quick to respond to this type of problem feedback.
This answer comes from the articleSpegel: using AI to transform web pages into an end-to-end browsing experienceThe
May not be reproduced without permission:AI productivity tools " How to debug the unsatisfactory Spegel conversion results?
English 
简体中文 
日本語 
Deutsch 
Português do Brasil