openclaw-soft-engine

Name: openclaw-soft-engine
Author: Syysean

Syysean March 10, 2026

2 copies 0 downloads

🦞 Industrial-grade Soft Execution Engine for OpenClaw. Features Multi-Expert Routing (DeepSeek-R1/V3), DAG Scheduling, API Key Pooling, and Docker Sandbox isolation.

Agent Definition

# 🦞 OpenClaw Soft Engine (Expert Matrix v2.0)

🌐 Language
- English (current)
- [简体中文](./README.zh-CN.md)

[![Docker](https://img.shields.io/badge/Docker-Ready-blue?logo=docker&logoColor=white)](https://www.docker.com)
[![Windows 11](https://img.shields.io/badge/OS-Windows_11-0078D4?logo=windows&logoColor=white)](https://www.microsoft.com/windows/windows-11)
[![Protocol](https://img.shields.io/badge/Protocol-DAG_Scheduling-orange)](workspace/AGENTS.md)
[![License](https://img.shields.io/badge/License-MIT-green)](LICENSE)
[![GitHub release](https://img.shields.io/github/v/release/Syysean/openclaw-soft-engine?color=orange&logo=github)](https://github.com/Syysean/openclaw-soft-engine/releases)
[![Last Commit](https://img.shields.io/github/last-commit/Syysean/openclaw-soft-engine?logo=git)](https://github.com/Syysean/openclaw-soft-engine/commits/main)

> **Note:** This project is a deployment template and enhancement layer for OpenClaw. It is not the official OpenClaw distribution.

OpenClaw Soft Engine is a production-oriented execution layer that upgrades OpenClaw with structured routing, multi-expert model orchestration, and hardened guardrails. It combines a DAG-based task protocol with multi-expert model routing to improve execution control, route-level isolation, and concurrent task handling.

This guide documents the evolution from v1.0 "Six-Model Matrix" to v2.0 "Soft Execution Engine."

## Why Soft Engine?

A comparison between a vanilla OpenClaw setup and the Soft Engine execution layer.

| Capability | Vanilla OpenClaw | Soft Engine v2.0 |
| :--- | :---: | :---: |
| **Execution Model** | Linear prompt flow | DAG-based execution (`PLAN` / `CHECKPOINT`) |
| **Model Routing** | Single-model or manual switching | Automatic multi-expert routing (Text / Vision / Reasoning / Code) |
| **Rate Limit Handling** | Shared API key (prone to 429 under concurrency) | Route-level API key isolation |
| **Memory Safety** | No built-in safeguards | Container memory limit (512

Comments

More Agents

View all

hybrid-model-workflow

HAL 分层混合模型工作流 — 强模型(Claude)负责理解/拆解/验收，低成本模型(DeepSeek)负责检索/提取/清洗。Hermes Agent skill。

ph4ble

Dynamic-Review-Agent

An LLM agent fine-tuned on DeepSeek for spaced repetition, dynamically integrating knowledge points based on the Ebbinghaus forgetting curve.

1838177

StellarOS-Watch

基于 STM32F103 构建的端到端 AI 智能手表生态。自研“零重定位”原生机器码动态加载引擎与页面栈式 UI 框架；集成生产级 OTA 回滚保护机制与高带宽（921600 baud）串口协议栈。通过 Node.js 中继实现 DeepSeek AI 语义控制及 ASRPRO 语音全双工交互，是一个集成了分布式计算、现代存储管理与 AI Agent 的嵌入式全栈工程。

chenshuang888