Linux Server — First Official Release
knowledge.db)shell_actions module. Running on macOS may work partially but is not officially supported.
/learn.txtserver.py — entry point, HTTP server + admin consoleengine.py — local Markov knowledge engineauth.py — user registration and loginchat_history.py / chat_loader.py — chat persistencefile_handler.py — file read/write (Linux paths)shell_actions.py — bash command runnergemini_client.py, groq_client.py, openrouter_client.py — AI backendspip install -r requirements.txt
.apikeys filekey_1: "AIza..." # Gemini key
key_2: YOUR_API_KEY_HERE # unused slot
key_3: YOUR_API_KEY_HERE # unused slot
# Localhost only (default, safest)
python server.py
# Custom port
python server.py --port 8080
# All interfaces (LAN/remote clients)
python server.py --public --port 5000
# No admin console (pure daemon mode)
python server.py --no-console
help to see available commands.
All POST bodies are JSON. All responses are JSON with an ok boolean.
POST /auth/register { username, password }
POST /auth/login { username, password }
POST /auth/logout { username }
POST /auth/change-password { username, current_password, new_password }
POST /auth/login-pin { pin }
POST /chat { username, message, model? }
GET /chat/history?username=<u>
POST /chat/load { username, pin }
POST /chat/clear { username }
POST /chat/export { username }
GET /files
POST /files/read { username, filename }
POST /files/read-path { username, filepath }
POST /files/write { username, filename, content }
POST /files/save-block { language, code, filename }
POST /files/clear-outputs {}
POST /ping { username?, model? }
GET /health
GET /models
GET /clock?username=<u>&model=<m>
| Command | What it does |
|---|---|
status | Knowledge stats + debug mode |
debug on/off | Toggle debug info in API responses |
learn <text> | Teach the local knowledge engine |
reset | Wipe and reseed knowledge base |
users | List registered accounts + PINs |
chats | List saved chats |
clear-outputs | Wipe the outputs/ folder |
models | List available AI models |
exit / quit | Stop the server |
| Model Name | Provider |
|---|---|
3.1-flash-lite default | Gemini |
3.1-flash | Gemini |
3.5-flash | Gemini |
llama-3.3-70b | Groq |
gpt-oss-20b | OpenRouter |
1.2
Linux Server
July 12, 2026
Stable
Linux Only
MIT
server.py
engine.py
auth.py
chat_history.py
chat_loader.py
saved_info.py
file_handler.py
shell_actions.py
code_extractor.py
markdown_highlighter.py
apikey_settings.py
gemini_client.py
groq_client.py
openrouter_client.py
requirements.txt
.apikeys
knowledge.db
users.env
users.pins
chats/
outputs/
saved-info/
.flask/user-settings/
.conf/server_config