AgentIndex icon
AgentIndex
ToolsMCP ServersHOTCategoriesTrendingNewBest ListsCompareGuides
Submit Tool
ToolsMCP ServersHOTCategoriesTrendingNewBest ListsCompareGuides
Home/
Dev Tooling/
sandbase-harness
sandbase-harness logo

sandbase-harness

Active·★ 635·Apache-2.0·Updated 2026-10-18
★ Trending

SandBase Harness is a local-first runtime designed for AI agents, providing essential infrastructure like persistent sessions, sandboxed execution environments, and comprehensive audit trails. It enables secure and observable agent operations directly on your machine or within your own infrastructure, supporting various models and tool integrations.

sandbase-harness sits in the Dev Tooling category, so it is usually better judged by workflow fit, deployment style, and team needs than by isolated features alone. Based on the available metadata, it leans most clearly toward Local, Docker, Kubernetes, and self-hosted worker sandboxes for safe code execution.. A representative use case is Running auditable coding agents with controlled and sandboxed code execution.. The listed license is Apache-2.0. GitHub stars are currently 640. Relevant tags include AI Agent Runtime, Sandboxed Execution.

This write-up is based on the project's README and public repository data. We have not hand-tested the tool.

#AI Agent Runtime#Sandboxed Execution#Session Management#Tool Orchestration#Audit Trails#Local-first Infrastructure#Multi-Model Support#Developer Console
$ Install
$ 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
↗ Visit site★ GitHub
01

Features

01Local, Docker, Kubernetes, and self-hosted worker sandboxes for safe code execution.
02Persistent sessions, resumable event streams, audit logs, and replay capabilities for inspecting long-running agents.
03MCP toolsets, credential vaults, and permission policies for controlling tool access.
04Compatibility with OpenAI, Anthropic, MiniMax, DeepSeek V4, and other OpenAI-compatible model providers.
05Local-first storage with SQLite for metadata and file storage, without a required hosted control plane.
02

Why choose it

+Local, Docker, Kubernetes, and self-hosted worker sandboxes for safe code execution.
+Persistent sessions, resumable event streams, audit logs, and replay capabilities for inspecting long-running agents.
+Running auditable coding agents with controlled and sandboxed code execution.
+Supports 7 platforms or runtime environments.
03

Trade-offs

!The final decision should still be made through side-by-side comparison with similar tools.
04

Compatibility

Node.js
Runtime
Verified via docs
npm
Package Manager
Verified via docs
Docker
Sandbox Backend
Verified via docs
Kubernetes
Sandbox Backend
Verified via docs
OpenAI API
Model Provider
Verified via docs
Anthropic API
Model Provider
Verified via docs
05

Quick start

1
$ git clone --branch v0.3.8 --depth 1 https://github.com/sandbaseai/sandbase-harness.git
2
$ cd sandbase-harness
3
$ npm ci
4
$ npm run build
5
$ mkdir ../my-agents
6
$ cd ../my-agents
7
$ node ../sandbase-harness/dist/index.js init
8
$ node ../sandbase-harness/dist/index.js start
06

Use cases

↳Running auditable coding agents with controlled and sandboxed code execution.
↳Using DeepSeek Harness as an interactive front end for agents, with rich session management.
↳Managing and securing third-party plugins with sandbox and filesystem protection.
↳Building an auditable research agent, combining evidence ledgers with sandboxed execution, credentials, audit, and replay.
↳Self-hosting an AI agent runtime on your own infrastructure for full control and data privacy.
07

Alternatives

⚡Top VS:
VScross-code-organizer★383→VSfastmcp★3.3k→VSCopilotKit★38.2k→VS1Panel★37.2k→
cross-code-organizer logo
cross-code-organizer★ 383
Cross-Code Organizer (formerly Claude Code Organizer): cross-harness config dashboard for Claude Code, Codex CLI, MCP servers, skills, memories, agents, sessions, security scanning, context budget, and backups.
VSCompare→
fastmcp logo
fastmcp★ 3.3k
A TypeScript framework for building MCP servers.
VSCompare→
CopilotKit logo
CopilotKit★ 38.2k
React UI + elegant infrastructure for AI Copilots, AI chatbots, and in-app AI agents. The Agentic Frontend.
VSCompare→
1Panel logo
1Panel★ 37.2k
🔥 1Panel provides an intuitive web interface and MCP Server to manage websites, files, containers, databases, and LLMs on a Linux server.
VSCompare→
agents-towards-production logo
agents-towards-production★ 21.7k
This repository delivers end-to-end, code-first tutorials covering every layer of production-grade GenAI agents, guiding you from spark to scale with proven patterns and reusable blueprints for real-world launches.
VSCompare→
unity-mcp logo
unity-mcp★ 15.3k
Unity MCP acts as a bridge, allowing AI assistants (like Claude, Cursor) to interact directly with your Unity Editor via a local MCP (Model Context Protocol) Client. Give your LLM tools to manage assets, control scenes, edit scripts, and automate tasks within Unity.
VSCompare→
fastapi_mcp logo
fastapi_mcp★ 12.0k
Expose your FastAPI endpoints as Model Context Protocol (MCP) tools, with Auth!
VSCompare→
nginx-ui logo
nginx-ui★ 11.7k
Yet another WebUI for Nginx
VSCompare→
See all alternatives →

Comments

Log in to leave a comment

No comments yet. Be the first!

On this page
01Features02Why choose it03Trade-offs04Compatibility05Quick start06Use cases07Alternatives
sandbaseai /
sandbase-harness
↗

Local-first, self-hosted AI agent runtime with MCP tools, sandboxed sessions, memory, credentials, audit/replay, and a local Console.

635Apache-2.0
Last commit5d ago
StatusActive
CategoryDev Tooling
Links
Documentation↗Discussion↗Issues↗Releases↗
Cloud Infrastructure
Ad

Deploy on DigitalOcean

Host your MCP servers, Docker containers, and AI agents with $200 free credit over 60 days.

Claim $200 Free Credit
© 2026 AgentIndex.app|Built by a 10-year iOS Developer.
Best by use caseAboutContactAdvertisePrivacy PolicyTerms of UseQYSGitHubBuy me a coffee ☕

Browse by Category

MCP ServersCode AssistantWorkflow AutomationRAG / Knowledge BaseMulti-AgentBrowser AutomationLLM InfraDev ToolingObservability

Not affiliated with Anthropic, OpenAI or Microsoft.