跳转至

sandbase-harness

⭐ 628 · ✅ 活跃 · 相关
类型 相关 分类 Harness
星数 ⭐ 628 状态 ✅ 活跃
作者 sandbaseai 更新时间 2026-08-20

一句话介绍

开源 CMA 兼容的任意模型 Agent 运行时：MCP 工具、沙箱会话、审计与回放。

详细介绍

English | 中文 AI-readable project metadata: llms.txt · installation guide A local-first runtime for AI agents. Sessions, sandboxed tools, memory, credentials, audit trails, and a built-in Console — all running on your machine or in your own infrastructure. git clone --branch v0.3.8 --depth 1 https://github.com/sandbaseai/sandbase-harness.git cd sandbase-harness npm ci npm run build mkdir ../my-agents && cd ../my-agents node ../sandbase-harness/dist/index.js init node ../sandbase-harness/dist/index.js start

✨ 核心特性

  • Claude Managed Agents-style /v1 API and local Console
  • SQLite-backed agents, sessions, environments, credential vaults, memory
  • local file/skill bytes stored in the workspace state directory
  • Resumable Server-Sent Events for session replay and debugging
  • One active model provider boundary configured through Settings V2
  • Sandbox backends: local process, Docker (per-session containers), Kubernetes
  • Settings V2: one workspace model vendor, loop engine, storage, memory,
  • MCP toolsets, permission policies, built-in tools, and skill packages

📦 安装

git clone --branch v0.3.8 --depth 1 https://github.com/sandbaseai/sandbase-harness.git
cd sandbase-harness
npm ci
npm run build
mkdir ../my-agents && cd ../my-agents
node ../sandbase-harness/dist/index.js init
node ../sandbase-harness/dist/index.js start
# open http://127.0.0.1:3000/dashboard

🚀 快速开始

node dist/index.js start --host 0.0.0.0

📚 更多信息

Quick Start

git clone --branch v0.3.8 --depth 1 https://github.com/sandbaseai/sandbase-harness.git cd sandbase-harness npm ci npm run build mkdir ../my-agents && cd ../my-agents node ../sandbase-harness/dist/index.js init node ../sandbase-harness/dist/index.js start Open http://127.0.0.1:3000/dashboard, go to Settings > Models, paste your API key, and you're running. The unscoped managed-agents name o

Configuration

.managed-agents/config.yaml: model: provider: openai api_key: ${OPENAI_API_KEY} storage: metadata: { provider: sqlite, options: {} } artifacts: { provider: local, options: { base_path: files } } Agents pick concrete model IDs (gpt-4o, claude-sonnet-4-20250514, openai/gpt-5.5). The workspace config only says how to reach the model service. For DeepSeek V4 Pro/Flash configuration, including

API Examples

Create an agent: curl -X POST http://127.0.0.1:3000/v1/agents \ -H "Content-Type: application/json" \ -d '{ "name": "Incident commander", "model": "gpt-4o", "system": "You are an on-call incident commander.", "tools": [{ "type": "agent_toolset_20260401" }] }' Create an environment (local sandbox): curl -X POST http://127.0.0.1:3000/v1/environments \ -H "Content-Type: application/json" \ -d '{ "nam

🔗 链接