-
Notifications
You must be signed in to change notification settings - Fork 0
st gen
b2o2i edited this page Apr 3, 2026
·
5 revisions
Sends your prompt file to an AI provider and saves the raw response into a .json container. This is the first step — the container it creates is used by every other command.
Run after: st-new · Run before: st-prep
st-gen subject.prompt # generate with default AI
st-gen --ai gemini subject.prompt # use a specific provider
st-gen --no-cache subject.prompt # bypass API cache
st-gen --prep subject.prompt # generate then run st-prepOptions: --ai --no-cache --prep --bang -v -q
Related: st-bang · st-prep · AI Providers
Writes a new entry to data[] in the container. Caching is MD5-keyed on the serialized request payload — two identical prompts to the same model always hit the cache. --bang N is used internally by st-bang (writes to tmp/ and creates a block file); don't call it directly.