[PR #48] [CLOSED] docs: add MateClaw integration guide #169

New issue

Closed

opened 2026-10-26 09:46:26 +00:00 by bsakaguchi · 0 comments

bsakaguchi commented

2026-10-26 09:46:26 +00:00

Owner

📋 Pull Request Information

Original PR: https://github.com/deepseek-ai/awesome-deepseek-agent/pull/48
Author: @mateaix
Created: 5/1/2026
Status: ❌ Closed

Base: main ← Head: add-mateclaw

📝 Commits (1)

e6ba263 docs: add MateClaw integration guide

📊 Changes

5 files changed (+178 additions, -0 deletions)

View changed files

📝 README.md (+1 -0)
📝 README.zh-CN.md (+1 -0)
➕ docs/assets/mateclaw_preview.png (+0 -0)
➕ docs/mateclaw.md (+88 -0)
➕ docs/mateclaw.zh-CN.md (+88 -0)

📄 Description

Summary

Adds an integration guide for MateClaw — an open-source personal AI operating system built on Spring Boot 3.5 + Spring AI Alibaba — covering DeepSeek V4 Pro and V4 Flash with thinking-mode reasoning.

New: docs/mateclaw.md (English) and docs/mateclaw.zh-CN.md (Simplified Chinese)
New: docs/assets/mateclaw_preview.png (dashboard screenshot)
Updated: README.md and README.zh-CN.md (table row added at the end)

What MateClaw is

One JAR. Apache 2.0. The same agent answers in the web console, the desktop app, the embeddable widget, the Java plugin SDK, and 8 IM channels (DingTalk, Feishu, WeChat Work, WeChat, Telegram, Discord, QQ, Slack). Same memory, same skills, same tools — different doors.

DeepSeek slots into a multi-vendor failover chain alongside DashScope, OpenAI, Anthropic, Gemini, Kimi, Ollama, LM Studio, MLX, and others. When the primary returns 401 or times out, the next provider picks up mid-sentence.

Why DeepSeek users may want it

Out of the box: deepseek-v4-pro and deepseek-v4-flash are pre-registered as builtin model entries (1M context, thinking-mode). Reasoning content streams alongside the answer and is preserved across the conversation.
ReAct + Plan-and-Execute agents on Spring AI Alibaba's StateGraph runtime.
LLM Wiki for knowledge digestion with per-sentence citations.
Memory lifecycle: post-conversation extraction, scheduled consolidation, dreaming workflows.
Skills + MCP (stdio / SSE / Streamable HTTP) + Tool Guard with RBAC and approval flow.

Test plan

Both new guide files render correctly (relative links, image path, language switcher header)
Both README updates land in the right table position with consistent column count
Image asset committed under docs/assets/

_{🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.}

## 📋 Pull Request Information **Original PR:** https://github.com/deepseek-ai/awesome-deepseek-agent/pull/48 **Author:** [@mateaix](https://github.com/mateaix) **Created:** 5/1/2026 **Status:** ❌ Closed **Base:** `main` ← **Head:** `add-mateclaw` --- ### 📝 Commits (1) - [`e6ba263`](https://github.com/deepseek-ai/awesome-deepseek-agent/commit/e6ba263c9d60dd4fcc9f518ee629f6ef2146962a) docs: add MateClaw integration guide ### 📊 Changes **5 files changed** (+178 additions, -0 deletions) <details> <summary>View changed files</summary> 📝 `README.md` (+1 -0) 📝 `README.zh-CN.md` (+1 -0) ➕ `docs/assets/mateclaw_preview.png` (+0 -0) ➕ `docs/mateclaw.md` (+88 -0) ➕ `docs/mateclaw.zh-CN.md` (+88 -0) </details> ### 📄 Description ## Summary Adds an integration guide for **MateClaw** — an open-source personal AI operating system built on Spring Boot 3.5 + Spring AI Alibaba — covering DeepSeek V4 Pro and V4 Flash with thinking-mode reasoning. - New: `docs/mateclaw.md` (English) and `docs/mateclaw.zh-CN.md` (Simplified Chinese) - New: `docs/assets/mateclaw_preview.png` (dashboard screenshot) - Updated: `README.md` and `README.zh-CN.md` (table row added at the end) ## What MateClaw is One JAR. Apache 2.0. The same agent answers in the web console, the desktop app, the embeddable widget, the Java plugin SDK, and 8 IM channels (DingTalk, Feishu, WeChat Work, WeChat, Telegram, Discord, QQ, Slack). Same memory, same skills, same tools — different doors. DeepSeek slots into a multi-vendor failover chain alongside DashScope, OpenAI, Anthropic, Gemini, Kimi, Ollama, LM Studio, MLX, and others. When the primary returns 401 or times out, the next provider picks up mid-sentence. ## Why DeepSeek users may want it - Out of the box: `deepseek-v4-pro` and `deepseek-v4-flash` are pre-registered as builtin model entries (1M context, thinking-mode). Reasoning content streams alongside the answer and is preserved across the conversation. - ReAct + Plan-and-Execute agents on Spring AI Alibaba's StateGraph runtime. - LLM Wiki for knowledge digestion with per-sentence citations. - Memory lifecycle: post-conversation extraction, scheduled consolidation, dreaming workflows. - Skills + MCP (stdio / SSE / Streamable HTTP) + Tool Guard with RBAC and approval flow. ## Links - Repo: https://github.com/matevip/mateclaw - Documentation: https://claw.mate.vip/docs - Live demo: https://claw-demo.mate.vip - Desktop downloads: https://github.com/matevip/mateclaw/releases ## Test plan - [x] Both new guide files render correctly (relative links, image path, language switcher header) - [x] Both README updates land in the right table position with consistent column count - [x] Image asset committed under `docs/assets/` --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>

bsakaguchi

2026-10-26 09:46:26 +00:00