- Add "get image" command - the LLM can now get an image from the bot's camera and analyze it (upon user request)
- Fix "love" LLM command - it wasn't seen by the LLM in the last version
- Cut out special characters from LLM response
- Under-the-hood intent matching change
- Include optional "requiresmatch" in intent data. If true for an intent, an exact match to the transcribed text is required. Otherwise, partial matches would work too. It is only true for "no" and "no thanks" at the moment.
Targets
- Windows (10/11, maybe 7)
- WirePodInstaller-v1.1.22.exe
- macOS (11 and above)
- WirePod-v1.1.22.dmg
- Android (4.4 and above)
- WirePod-1.1.22.apk
- Debian/Ubuntu (Debian Jessie> and Ubuntu 16.04>)
- wirepod_amd64-v1.1.22.deb (most computers)
- wirepod_armhf-v1.1.22.deb (SBCs like Raspberry Pi with either 32-bit or 64-bit OS)
- wirepod_arm64-v1.1.22.deb (SBCs with 64-bit OS. Required for Raspberry Pi 5.)