diff --git a/README.en.md b/README.en.md
index fc2caed95..91d295b56 100644
--- a/README.en.md
+++ b/README.en.md
@@ -314,7 +314,7 @@ DeepSeek Harness is DeepSeek's open-source agent harness — a runnable coding a
 - [mjylfz/dsh-skill-mover](https://github.com/mjylfz/dsh-skill-mover) - One-click migration of skills from other agents into DeepSeek Harness: 14 platform detectors plus shared-layer recognition, same-name skill merging, no re-copying of existing skills, and rollback support.
 - [moon09300731/dsh-approval-gate](https://github.com/moon09300731/dsh-approval-gate) — Risk-gated approval automation: flash pre-classifies whether a write/command is irreversible — safe operations are auto-approved, dangerous ones are escalated to human approval (fail-safe).
 - [moon09300731/dsh-vision-tools](https://github.com/moon09300731/dsh-vision-tools) — Full vision-capability bundle for DeepSeek Harness: a `vision_understand` tool (OpenAI-compatible vision APIs, free Zhipu GLM-4V-Flash by default) plus paste/drag-and-drop/button entry points for image recognition.
-- [Nicholas023/vision-exp-tile](https://github.com/Nicholas023/vision-exp-tile) - Large-image recognition for vision models: 800×800 lossless tiling (smart/pipeline/full strategies) plus local OCR (auto inversion/binarization/handwriting split) and optional multi-vendor GPU acceleration (DirectML covering NVIDIA/AMD/Intel, CUDA/OpenVINO) with automatic CPU fallback; MIT licensed (111 unit tests).
+- [Nicholaskin/vision-exp-tile](https://github.com/Nicholaskin/vision-exp-tile) - Large-image recognition for vision models: 800×800 lossless tiling (smart/pipeline/full strategies) plus local OCR (auto inversion/binarization/handwriting split) and optional multi-vendor GPU acceleration (DirectML covering NVIDIA/AMD/Intel, CUDA/OpenVINO) with automatic CPU fallback; MIT licensed (111 unit tests).
 - [nousag/dsh-writing-desk](https://github.com/nousag/dsh-writing-desk) - Local-first writing workbench: book/volume/chapter management, draft editor, note reference, word count and export.
 - [nvbb/dsh-ollama-vision-bridge](https://github.com/nvbb/dsh-ollama-vision-bridge) - When the chat model is text-only, describe attached images with a local Ollama VL model (qwen3-vl:8b) and inject into the same model step: local vision, no cloud (DSH ≥ 0.1.2-rc.1).
 - [omdsh-dev/dsh-custom-tool](https://github.com/omdsh-dev/dsh-custom-tool) - Create and manage sandboxed JavaScript tools with a Monaco editor.
diff --git a/README.md b/README.md
index 740be2433..42c08f63b 100644
--- a/README.md
+++ b/README.md
@@ -314,7 +314,7 @@ DeepSeek Harness 是 DeepSeek 开源的 agent harness——既是可直接运行
 - [mjylfz/dsh-skill-mover](https://github.com/mjylfz/dsh-skill-mover) — 把其他 Agent 的技能一键搬进 DeepSeek Harness：支持 14 个平台 + 共享层识别、同名技能合并、不重复搬运、可回滚。
 - [moon09300731/dsh-approval-gate](https://github.com/moon09300731/dsh-approval-gate) — 自动审批门控：Flash 预判写入/命令是否不可回补，安全操作自动批准、危险操作转人工（fail-safe）。
 - [moon09300731/dsh-vision-tools](https://github.com/moon09300731/dsh-vision-tools) — DeepSeek Harness 视觉能力全家桶：vision_understand 工具（OpenAI 兼容视觉 API，默认免费智谱 GLM-4V-Flash）+ 粘贴/拖拽/按钮三入口识图。
-- [Nicholas023/vision-exp-tile](https://github.com/Nicholas023/vision-exp-tile) — 面向视觉大模型的大图智能识图：800×800 无损切块识别（smart/pipeline/full 三策略）+ 本地 OCR（自动反色/二值化/手写分流）+ 可选多厂商 GPU 加速（DirectML 覆盖 NVIDIA/AMD/Intel，CUDA/OpenVINO），失败自动回退 CPU；MIT 开源（111 项单元测试）。
+- [Nicholaskin/vision-exp-tile](https://github.com/Nicholaskin/vision-exp-tile) — 面向视觉大模型的大图智能识图：800×800 无损切块识别（smart/pipeline/full 三策略）+ 本地 OCR（自动反色/二值化/手写分流）+ 可选多厂商 GPU 加速（DirectML 覆盖 NVIDIA/AMD/Intel，CUDA/OpenVINO），失败自动回退 CPU；MIT 开源（111 项单元测试）。
 - [nousag/dsh-writing-desk](https://github.com/nousag/dsh-writing-desk) — 本地优先写作工作台：书籍/卷/章管理、正文编辑、笔记资料调用、字数统计与导出。
 - [nvbb/dsh-ollama-vision-bridge](https://github.com/nvbb/dsh-ollama-vision-bridge) — 聊天选中纯文本模型时，自动用本地 Ollama VL 模型（qwen3-vl:8b）描述附图并注入同一模型步骤：本地视觉、显存冷却、无需云端（DSH ≥ 0.1.2-rc.1）。
 - [omdsh-dev/dsh-custom-tool](https://github.com/omdsh-dev/dsh-custom-tool) — 用 Monaco 编辑器创建和管理沙箱化的自定义 JavaScript 工具。