Use cases

MOB³ is a terminal browser with a headless MCP mode, so it fits anywhere a graphical browser does not: SSH sessions, containers, CI runners, remote servers and AI agents. It is small and text-focused, not a full browser engine, so the best results come from content pages, docs and server-rendered HTML; see site support.

Remember that MOB³ is not a secure browser: no sandboxing, no site isolation, no CSP. Use it for reading and automation, never for passwords, accounts, payments or confidential data. See security.

Developers and sysadmins

• Terminal browsing — browse websites directly from SSH sessions, servers, containers and minimal Linux environments.
• Headless website testing — load pages, execute JavaScript, inspect rendered content and verify website behavior without Chromium.
• Web scraping — extract text, links, metadata and other content from JavaScript-driven pages.
• Website debugging — inspect how HTML, CSS and JavaScript render without opening a graphical browser.
• Screenshot generation — capture web pages as PNG images directly from scripts or the command line (mob3 -png ... > page.png).
• CI/CD integration — include lightweight browser checks in automated pipelines.
• Server-side automation — automate web interactions on remote machines without a desktop environment.

AI agents and MCP

Start with mob3 -mcp and the automation reference.

• AI web browsing — give AI agents the ability to browse and interact with real websites through MCP.
• AI research assistants — let agents navigate websites, follow links, collect information and summarize findings.
• Autonomous web automation — allow agents to fill out forms, click buttons, navigate pages and execute workflows.
• Local AI browser — provide a browser that runs locally, without sending browser activity through a third-party automation service.
• Agentic workflows — integrate web browsing into multi-step AI workflows and scripts.
• Website inspection by AI — let agents inspect rendered pages, discover interactive elements and understand page structure.

Everyday terminal users

• SSH browsing — browse the web from remote servers without installing a desktop environment.
• Distraction-free browsing — access websites without the overhead of a graphical browser.
• Low-resource environments — browse on machines with limited memory, CPU or graphical capabilities.
• Privacy-conscious browsing — keep browsing history, cookies, storage and caches in memory only; nothing is written to disk.
• Inline image viewing — view web images directly in compatible terminals with -graphics=auto|kitty|iterm|blocks.

Automation and integration

• Shell scripting — fetch and interact with websites as part of shell scripts, for example with mob3 -dump URL.
• Monitoring — periodically inspect web pages and detect content changes.
• Documentation lookup — browse online documentation directly from development environments.
• Link checking — visit URLs and inspect whether pages render correctly.
• Content conversion — turn rendered web pages into text or screenshots for further processing.
• Custom tooling — build other applications and agents on top of MOB³'s browser capabilities.