This is most likely caused by the injector trying to inject "uninjectable" headers (e.g. te or accept-encoding..., which all depend on the browser capabilities and cannot be easily "faked").
We should implement some filter to only inject "safe" headers with user-agent data, not headers that
depend on the requested content.
depend on the browser capabilities (supported encodings, )
barjin
commented
5 months ago
Reported on Slack:
This is most likely caused by the injector trying to inject "uninjectable" headers (e.g.
teoraccept-encoding..., which all depend on the browser capabilities and cannot be easily "faked").We should implement some filter to only inject "safe" headers with user-agent data, not headers that