Repository navigation
samagotchi 0.95.0
Removed
- chi's own
mlxandomlxtransports (api:/transport:on a host,server.transport,
SAMAGOTCHI_SERVER_TRANSPORT): mlx_lm.server and oMLX are chat hosts now,url: http://HOST:PORT/v1, api: openai
(on oMLX the chat endpoint matched or beat the native one, tool calls intact). llama.cpp is the only raw-prompt
server. A config that still namesmlxoromlxstops chi at start with one line:Error: config.yml hosts.mac.api: omlx was removed; for mlx_lm.server or oMLX use url: http://HOST:PORT/v1, api: openai in a hosts: entry. Name the model as the server's/v1/modelslists it (oMLX's prefixedmlx-community--…id); the short
selector oMLX's transport used to resolve now 404s. See
docs/configuration.md. - The transport setting:
server.transport, a host'stransport:,SAMAGOTCHI_SERVER_TRANSPORTand
--server-transport(llama.cpp is the one raw-prompt server, and a host withoutapi:talks it, as
api: llama_cpp). A config or environment still setting one tollama_cppwarns at start and runs on:
Warning: config: 'server.transport' was removed: llama.cpp is chi's only raw-prompt transport (a hosts entry without api: is api: llama_cpp); remove it;mlxoromlxstops chi as above. The flag is an invalid option. - The Qwen turn preamble: a native
qwen36prompt no longer asks the model to open its thinking with aTURN:line,
and thethinking.turn_preamblesetting goes with it (--no-thinking-turn-preamble,
SAMAGOTCHI_THINKING_TURN_PREAMBLE). A config.yml that still sets it getsWarning: config: unknown key 'thinking.turn_preamble'at start and runs on; the flag is an invalid option.
Fixed
- Native Qwen3.6 on llama.cpp (
api: llama_cpp, profileqwen36) thinks again: where the server's chat template
opens a thought after the assistant cue (Qwen3.5 and later, read from the model's/props), chi's prompt ends with
<think>and a newline as the template's does, at every level butoff; before, the model wrote an empty thought
and answered at once. Its reasoning streams as thinking and stays out of the answer. Turns take longer;thinking: offanswers at once as before. An older Qwen3 or other template, or no template in/props, keeps the old prompt.
low,mediumandhighthink the same there, which the notice now says. A tool call the model writes inside a
closed thought no longer runs. See docs/configuration.md. chi bundle trywith a config chi stops at (e.g. a removedapi: omlx) fails the install with that one line, not a
backtrace.
Update with chi update (the system bundle 0.95.0), then restart chi web.