@deep-tui/plugin-usage-inline
v0.1.0
Published
Per-message token and cost annotations for Deep TUI's TUI.
Readme
@deep-tui/plugin-usage-inline
Appends a compact usage line to every completed assistant response:
↳ cost $0.000035 · tok in 100 · out 24 · cache 40The TUI footer simultaneously shows the running session cost. Per-message
usage is stored with durable assistant records, so the
annotations return when a conversation is resumed. Set inline: false or
footer: false to enable only one surface; costPrecision accepts 2–10.
The visible label is simply “cost.” Token counts come from the provider response and the amount uses the configured provider pricing. It is not represented as a per-request invoice when the provider does not expose one.
