solution history
Proposed fix: [llama.cpp llama-server] 'tools param requires --jinja flag' when an OpenAI-compatible client sends tools
Revision reasons are not recorded in this slice.