summaryrefslogtreecommitdiff
path: root/docs
diff options
context:
space:
mode:
authorPaul Buetow <paul@buetow.org>2026-03-14 09:21:05 +0200
committerPaul Buetow <paul@buetow.org>2026-03-14 09:21:05 +0200
commit4137c95d0b6882fbdc40d73b1f3528c28ceb653a (patch)
treeb8c1f53c2d7e657e8bca59b6154fefd128b1fb30 /docs
parent8ae4be9684a58d44985e5b5ee5e90f74555b2dde (diff)
Release v0.22.1v0.22.1
Diffstat (limited to 'docs')
-rw-r--r--docs/usage.md2
1 files changed, 2 insertions, 0 deletions
diff --git a/docs/usage.md b/docs/usage.md
index e7b59a8..1ecf30f 100644
--- a/docs/usage.md
+++ b/docs/usage.md
@@ -85,6 +85,8 @@ Process text via the configured LLM:
Defaults: concise answers. If the prompt asks for commands, Hexai outputs only commands. Add the word `explain` to request a verbose explanation. Exit codes: `0` success, `1` provider/config error, `2` no input`.
+Provider/model headers and run summaries are written to `stderr`, so `stdout` stays usable in pipes.
+
`--tps-simulation` accepts either a fixed rate such as `20` or a range such as `12-18`. It streams positional arguments, piped stdin, or built-in placeholder text when no input is provided, so you can preview perceived model latency without needing a real provider or local hardware.
### Examples