3 Codex Skill Ocr Vision DSH Plugins
A cluster of 3 repos centred on codex skill, claude code skill and ocr vision.
How this collection was built
This collection is machine-derived: DeepSeek Harness plugin READMEs are vectorised with TF-IDF and grouped by cosine similarity, so membership is a statement about wording, not an endorsement — it can be, and sometimes is, overridden by hand. No official DSH extension point maps cleanly onto this cluster; browse categories for the authoritative axis. Members are ordered by GitHub stars. Clustered on 2026-08-15.
Members, most-starred first
Image Vision
wangyang10
No description provided.
DeepSeek Prism
YOGEMOW
为纯文本模型按需识图：DSH 零补丁 Cordis 插件（prism_see 工具 + 图片 VEP 降级 + 技能运行时注册）+ Codex Skill；多 Provider 视觉 API，VEP/1 低 Token 视觉证据包
Delegate To Deepseek Harness
papperrollinggery
Local-first Codex Skill for scoped DeepSeek Harness delegation across copywriting, video pre-production, coding, and review.