FAQ - Frequently Asked Questions¶
Why does my filter not trigger?¶
- Filters are matched as whole words. Use parentheses to group multiple triggers:
Note
Patrizio uses case-insensitive matching. Finally, make sure you're not using a different character set!
More can be found in its filter usage guide
What if I want to add a new command?¶
Please refer to the Developer documentation for further information about Patrizio's architecture.
How to run tests locally?¶
Please refer to the project README for further information.
Is there a public instance?¶
Not yet, due to data being saved in plain without any security. I'll think about a public instance once some sort of privacy measure has been taken in that sense
Why does /prompt say it's not configured?¶
The /prompt command requires an OpenAI-compatible API key. If the bot operator hasn't set openai_api_key in the
configuration, the feature is disabled. Check the configuration docs for setup instructions.
Why does /prompt say I'm not authorized?¶
The bot operator can restrict which chats are allowed to use the AI feature via the openai_allowed_chat_ids setting.
If your chat is not in the list, you'll receive a "not authorized" message. Ask the person running the bot to add your
chat ID to the allowlist.
Can I use a local AI model instead of OpenAI?¶
Yes! Patrizio supports any OpenAI-compatible API. You can point it at Ollama,
LMStudio, or any other provider by setting openai_base_url to the local endpoint. For
example, for Ollama you would set openai_base_url = "http://localhost:11434/v1".
How do I start a fresh AI conversation?¶
Each /prompt creates an independent thread. To start over, simply send a new /prompt message instead of replying to
an existing conversation chain.