Suchintan
Suchintan
> i think 90% people can't run the program.... :'(
@ryanholtschneider2 that's interesting -- we have a check to see if postgres is already running before running commands. https://github.com/Skyvern-AI/skyvern/blob/main/setup.sh#L179 A few questions: 1. Did you have it running in Docker?...
What kind of documentation are you looking for? We have API documentation here: docs.skyvern.com
Thanks for the feedback here -- I'll add it to our roadmap!
@suyashdb I think this is the problem: ``` - VITE_WSS_BASE_URL=ws://localhost:8000/api/v1 ``` I think we would expect this url to be pointing to the image running the Skyvern instance (not UI...
Referencing localhost (or 0.0.0.0) will work if all containers are running on the same node
Hey! We already have this feature :) You can grab all of the HAR logs of a particular Skyvern session out as an artifact, which contains all network calls I...
We do store the network requests as HAR artifacts and can be found under here: `export HAR_PATH=/data/har`
[https://github.com/Skyvern-AI/skyvern/blob/3f209404f7b72c223c54ed55d9e93e5e4d04cb26/skyvern/forge/agent.py#L1593](https://github.com/Skyvern-AI/skyvern/blob/3f209404f7b72c223c54ed55d9e93e5e4d04cb26/skyvern/forge/agent.py#L1593)You can retrieve it from the step endpoint by querying the final step id
The HAR file contains all network traffic (not related to LLMs)