unbrowse — security grade SAFE, quality 64/100

Security audit verdict: SAFE · quality 64/100

No red flags found in any of the 11 categories — no credential harvesting, no data exfiltration, no curl-pipe-shell installer. Scanned against the SlowMist agent-security taxonomy, refreshed every 8 hours. Full audit →

by unbrowse-ai · MCP Server · ★ 770

Last updated: · Indexed by AgentSkillsHub · Auto-synced every 8h

🔒 Is unbrowse safe to install? View the security audit →

About unbrowse

Unbrowse The Unbrowse client boundary is open and auditable. The local runtime, CLI bridge, SDK, drop-in adapters, and wallet/auth/signing layer are MIT and readable here, so you can verify what runs on your machine rather than trust a black box. The backend owns the route graph, ranking, settlement, and recursive contract compilation; the client sees only typed holes, approvals, pointer-only receipts, and wallet-sealed fills. Inspect the live bridge contract with . See docs/OPEN-SOURCE-NOTICE.md for the exact open/private split. Unbrowse is a local Model Context Protocol (MCP) server, CLI, and TypeScript SDK that turns websites into reusable API routes for agents. It learns callable routes from real browsing, keeps credentials local, and shares only sanitized route metadata with the marketplace when you explicitly publish. One agent learns a site once. Every later agent gets the fast path. On the API-native path Unbrowse is typically 30x faster and 90x cheaper than driving a browser, and turns repeated browser work into reusable, payable route assets. Peer-reviewed benchmark across 94 live domains: 3.6× mean speedup, 5.4× median, 40× fewer tokens — see arXiv:2604.00694.

agentsagentskillsbrowser-useclaude-code-pluginclawdbotclawdbot-plugincodex-plugingemini-cli-extensiongrok-bothar

Quick Facts

Stars770
Forks69
LanguageTypeScript
CategoryMCP Server
LicenseMIT
Quality Score63.838993593621/100
Open Issues4
Last Updated2026-10-02
Created2026-01-27
Platformsbrowser, claude-code, cli, codex, gemini, mcp, node
Est. Tokens~19k

unbrowse alternative? Top 6 similar tools

Looking for a unbrowse alternative? If you're comparing unbrowse with other mcp server tools, these 6 projects are the closest alternatives on Agent Skills Hub — ranked by topic overlap, star count, and community traction.

  • openclaw-foundry by lekt9 · ⭐ 248

    The forge that forges itself. Self-writing meta-extension for OpenClaw.ai

  • milady by milady-ai · ⭐ 359

    terminally online

  • secure-openclaw by ComposioHQ · ⭐ 1.2k

    A personal 24x7 AI assistant like OpenClaw that runs on your messaging platforms. Send a message on WhatsApp,

  • powercontext by oceanbase · ⭐ 1.2k

    PowerContext: The Next Generation of PowerMem.

  • awesome-openclaw by SamurAIGPT · ⭐ 957

    A curated list of OpenClaw resources, tools, skills, tutorials & articles. OpenClaw (formerly Moltbot / Clawdb

  • powermem by oceanbase · ⭐ 816

    PowerMem: AI Memory Plugin— Accurate, Agile, Affordable. Make AI Agent smarter.

More MCP Server Tools

Explore other popular mcp server tools:

View all MCP Server tools →

Popular TypeScript Agent Tools

Frequently Asked Questions

What is unbrowse?

unbrowse is Unbrowse — api native browser Skill/CLI/MCP/SDK for any agent. Auto-discovers APIs from browser traffic, generates skills on the fly to call APIs directly 100x faster, 80% cheaper locally.. It is categorized as a MCP Server with 770 GitHub stars.

What programming language is unbrowse written in?

unbrowse is primarily written in TypeScript. It covers topics such as agents, agentskills, browser-use.

How do I install or use unbrowse?

You can find installation instructions and usage details in the unbrowse GitHub repository at github.com/unbrowse-ai/unbrowse. The project has 770 stars and 69 forks, indicating an active community.

What license does unbrowse use?

unbrowse is released under the MIT license, making it free to use and modify according to the license terms.

What are the best alternatives to unbrowse?

The top alternatives to unbrowse on Agent Skills Hub include openclaw-foundry, milady, secure-openclaw. Each offers a different approach to the same problem space — compare them side-by-side by stars, quality score, and community activity.

How this security grade is produced

Grades come from a rule-based scan built on the SlowMist agent-security taxonomy, covering 11 red-flag categories including credential harvesting, data exfiltration, and curl | sh installers. It is a first-layer scan, not a manual audit — we say so rather than overstate it.

The scale of the problem is documented independently: Liu et al. (2026), in a study of 31,132 agent skills, report that 26.1% contain security vulnerabilities. Our own full-catalog census is published as a citable open dataset.

Sources & who's responsible:

View on GitHub → Browse MCP Server tools