indifferentketchup a8e475fdf4 perf(llama): unshadow cache-type + spec-decoding flags for agent opt-in ...

KV cache quantization (--cache-type-k q4_0) and ngram speculative decoding
(--spec-type ngram-mod) are high-value llama.cpp features that improve VRAM
usage and tokens/sec. Removing them from the shadowing lists allows agents
to enable them via llama_extra_args.

2026-06-07 22:40:23 +00:00

..

booterm

refactor: codebase audit cleanup — dead code, dedup, module splits

2026-06-02 21:12:29 +00:00

coder

feat(coder,server): audit engine — session audit, guideline compliance, user correction tracking

2026-06-07 22:16:35 +00:00

server

perf(llama): unshadow cache-type + spec-decoding flags for agent opt-in

2026-06-07 22:40:23 +00:00

web

feat(web,server): inference settings UI with per-session inference overrides

2026-06-07 22:16:29 +00:00