← 返回列表
D

@goodandready/dsh-vision-bridge

v0.1.1 · 0 天前发布

Self-owned vision bridge for DeepSeek Harness: pick the vision model yourself from the available ones. Image content blocks are rewritten into text markers at the agent boundary so text-only models never fail a turn, and a describe_image tool answers via

deepseek-harnessdshdsh-pluginvisionvision-modelimagemultimodaltext-to-image

README

dsh-vision-bridge

Vision bridge for DeepSeek Harness (dsh) — a self-contained replacement for dsh-vision-router.

When the chat model has no vision (e.g. deepseek-v4-flash) and a message contains an image, the image never reaches the text-only model. Instead, the plugin substitutes an automatic text description from the vision model you choose (Hermes-style). The text model just reads text and keeps the conversation going; the image stays in the session log and in the UI.

  • describe_image — a tool for when you need more precise details on an image (ask a model).

  • Description cache by contentHash of the bytes — an image is described once, then the previous description is reused on later turns.

  • **You pick t
    • 查看完整 README →

    版本兼容矩阵

    DSH 版本状态说明
    dsh@0.1.0-rc.6 ✅ 兼容 cordis patch ✅
    dsh@0.1.0 ✅ 兼容 标准 patch ✅
    dsh@0.2.0 ⚠️ 待验证 待实测（API 可能在变）
    dsh@0.3.0+ ❓ 未知 尚未发布

    💡 兼容矩阵每周末自动跑实测，欢迎 PR 修正

    元数据

    License
    MIT
    npm
    @goodandready/dsh-vision-bridge
    Repo
    git+https://github.com/GooDAnDReaDY/dsh-vision-bridge.git
    Engines
    {"node":">=20"}
    Weekly DL
    0
    Monthly DL
    0
    dsh.bundle
    {"bundle":{"patch":"./cordis.patch.yml"},"client":{"platform":"web","inject":["@deepseek-ai/dsh-client-runtime","@deepseek-ai/dsh-client-locale","@deepseek-ai/dsh-client-ui-slots","@deepseek-ai/dsh-client-ui-settings"]}}