ethanweave/dsh-vision-better
Gives a text-only model in DeepSeek Harness the ability to work with images without switching model or session: if the active model cannot accept an image, the plugin relays it to your own GLM-compatible vision endpoint and hands the analysis to the original model as hidden context — no new buttons, modes or assistant bubbles, image and provider/model details never touched.
dsh-vision-better gives a text-only model in DeepSeek Harness the ability to work with images without switching the model, starting a new session, or changing a visible message. You paste an image into the chat as usual; if the active model cannot accept images, the plugin sends the image to your own GLM-compatible vision endpoint and hands the analysis to the original model as hidden context, so the original model keeps answering in the same session with its history intact. There are no new chat buttons, modes, model entries, assistant bubbles or extra messages, and no vision request at all when there is no image or when the active model already accepts images. The API key is stored by DSH's credential service on the host and is never sent back to the browser; the image, your text, the session id and the selected provider/model are never touched.
Install
dsh plugin --profile web add github:ethanweave/dsh-vision-betterCopied verbatim from the README's Install block ('from GitHub'); the same block shows a released-tarball URL and a local-clone path. The README states the plugin has zero dependencies and no build step, and that the separate core patch is required for image admission.
Compatibility
Requires DeepSeek Harness 0.2.0-rc.2 or newer within 0.2.x, Node >= 22.19, and a Zhipu (智谱) open-platform API key or any GLM-compatible vision endpoint. The README is explicit that stock DSH rejects an image before the message is stored when the active model does not declare image support, and that no public plugin hook can influence that check — so this repository also ships a small core patch adding a session/image-admission waterfall. Without the extension the hidden-vision-context half still works (nothing breaks) but images never reach a text-only session.
Details
- Repo: ethanweave/dsh-vision-better
- Category: Vision & Multimodal
- Stars: 1
- Version: GitHub repo ethanweave/dsh-vision-better (install is a github: spec to the repo itself).
Alternatives
54xkeee/dsh-vision · 54xkeee/dsh-youreyes · Cheng-cheng9669/dsh-deepseek-vision