vuvanmai936-dot/dsh-ocr-bridge ↗★ 0
dsh-ocr-bridge
将图片粘贴至 DeepSeek Harness 聊天中，并在纯文本 DeepSeek 模型回答前，通过免费的本地后端（macOS Vision / Tesseract）读取图像内容。
AI 分析
核心用途是为纯文本模型补充本地 OCR 识图能力。适合需要粘贴图片并进行本地免费文字识别的用户。要求 DSH 0.1.0-rc.7。
安装
$
npx -p @deepseek-ai/dsh dsh plugin --profile web add github:vuvanmai936-dot/dsh-ocr-bridge说明文档
阅读完整 README ↗Configuration
All settings are optional. They live in the llm-deepseek settings section (the official DeepSeek section this plugin takes over), so your existing DeepSeek settings keep working.
settings.yaml (or the GUI settings surface):
llm-deepseek:
visionTimeoutMs: 180000 # per-backend recognition timeout (ms), default 180000
maxImages: 8 # images recognized per request, 1..32, default 8
# …all official DeepSeek settings keep working: baseURL, apiKeyEnv, thinking, …
Notes:
- Over the limit, the request fails with
VISION_IMAGE_LIMITinstead of silently dropping images. - Recognized observations are cached per (image set + latest user text), up to 64 entries, so re-sending the same screenshot in one conversation does not re-run OCR.
DEEPSEEK_API_KEYresolution is fully preserved:ctx.credentialsfirst, then the launch environment, exactly like the official adapter.- png / jpeg / webp / gif are all supported (whatever the harness attachment admission accepts).