From 8dbbe7a76d002566b8be29332a9e5bef8789dc1a Mon Sep 17 00:00:00 2001
From: Green Sky <green@g-s.xyz>
Date: Tue, 6 Feb 2024 14:07:02 +0100
Subject: [PATCH] add small docu to readme

---
 README.md | 47 +++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 47 insertions(+)

diff --git a/README.md b/README.md
index 11a2dd1..308a380 100644
--- a/README.md
+++ b/README.md
@@ -1 +1,48 @@
+# llama.cpp webserver based api + RPBot
+
 used llama embedded webserver api for easier portability
+
+## LlamaCppWeb config options
+
+- `LlamaCppWeb`, `server`
+  - type: `string`
+  - host + port
+  - default: `localhost:8080`
+
+
+## RPBot config options
+TODO: move rpbot to own repo
+
+- `RPBot`, `system_prompt`(, opt contact ID)
+  - type: `string`
+  - System prompt that is prefixed
+  - can contain spcific formatters
+    - `{self_name}` username for specified chat
+  - default: `Transcript of a group chat, where {self_name} talks to online strangers.\n{self_name} is creative and curious. {self_name} is writing with precision, but also with occasional typos.\n`
+
+- `RPBot`, `min_messages`(, opt contact ID)
+  - type: `int`
+  - TODO: implement per ID
+  - default: `4`
+
+- `RPBot`, `max_interactive_delay`(, opt contact ID)
+  - type: `float`
+  - TODO: implement per ID
+  - default: `4.0`
+
+
+## example config
+
+```json
+{
+	"LlamaCppWeb": {
+		"server": "localhost:8080"
+	},
+	"RPBot": {
+		"min_messages": 1,
+		"system_prompt":  "A group chat with an artificial intelligence assistant called {self_name}. The assistant gives helpful and detailed answers to the chat participants's questions. The assistant has it's own opinions. {self_name} only ever answers when explicitly asked.\n"
+		}
+	}
+}
+```
+