#627 · Primary category: AI Agents & Automation

mateclaw

agent agent-harness ai-agent dingtalk-robot feishu-bot llm-wiki loop-engineering mcp-protocol multi-agent multiagent openclaw plan-and-execute skills spring-ai spring-boot telegrambot tool-calling

🤖 MateClaw — Your second brain with Multi-Agent Orchestration, MCP Protocol, Skills & Memory, Dream, and Multi-Channel Support. Built on Spring AI Alibaba.

Project last updated:08/29/26

GitHub Stars

1.1K

Forks

320

Contributors

19

License

Apache-2.0

Why we included this project

MateClaw is a genuinely deployable multi-agent runtime rather than a demo scaffold, which is why teams evaluating agent infrastructure should look at it. You get a Java/Spring Boot harness that runs as a single JAR and coordinates several "digital employees," each with a role, goal, and tools, using ReAct and Plan-and-Execute on a state-graph runtime. It stands out because the agent loop, knowledge base, memory, skills, and chat channels are built together rather than bolted on, so knowledge, memory, tools, and reasoning all live in one deployment. That matters for organizations that want one auditable system their operations team can actually sign off on, with multi-user workspaces, approval gates on sensitive tool calls, a full audit trail, and provider failover when a model goes down. Teams on Slack, Feishu, DingTalk, Telegram, or Discord get the same agent behavior across channels while the web console, desktop app, and embeddable widget act as additional access points.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category