Multimodal deepseek harness plug-in with built-in model
linxuhao/Deepseek-Continuity · skills
A multimodal deepseek harness plug-in with its own model, providing generated image/sound consistency.
★ 2GitHub stars
0Forks
2026-08-24Last updated
PythonLanguage
MITLicense
Key features
- What the reload actually costs4.8 s instead of 1.2 s, and only on the first call after switching away.
- Generation backends are stateless: ask for the same character twice and you get two people who merely resemble each other.
- A backend that miscomputes returns a perfectly well-formed all-zero WAV, or a flat grey PNG, with HTTP 200.
- Do not regex the prose for paths.
- Ten dialogue lines in a row pay it once
Requirements
- Audio-only installs (see below) need 20 GiB during install, 9.5 GiB after.
- All VRAM/RAM figures on this page are GiB (2³⁰ bytes), which is what rocm-smi and vulkaninfo report.
Install command
dsh plugin --profile web add github:linxuhao/Deepseek-Continuity