From 6ae195baa09bc217d51f3fb0a38312dfdcc85fdb Mon Sep 17 00:00:00 2001 From: "Danilo M." Date: Tue, 25 Aug 2026 18:42:40 +0200 Subject: fix: honor skills_enabled on restore, and final-review polish --- README.md | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'README.md') diff --git a/README.md b/README.md index 77ce895..349c21d 100644 --- a/README.md +++ b/README.md @@ -495,6 +495,11 @@ the whole session in chat mode and drop after the reply in one-shot mode. `max_searches` caps the total tool rounds per turn here too, so skills load and searches share the same per-turn budget. +Skill text is injected verbatim into the system prompt. Skills are +user-authored files in a directory you control, trusted the same way the +prompts in `prompts/` are, but treat a skill downloaded from elsewhere like +any other untrusted instructions. + ### External providers llamachat can talk to OpenAI-compatible cloud providers alongside the local -- cgit v1.2.3