AI Agent Hub
Back to plugins
🧰

dsh-tui-browser-use

Web Tools Updated 2026.09.04

Run the following command in DeepSeek Harness:

dsh plugin install FlameTN7/dsh-tui-browser-use

Paste the following prompt into your AI chat to install this plugin:

Run dsh plugin install FlameTN7/dsh-tui-browser-use in DeepSeek Harness to add this browser automation plugin; the source repository is available at https://github.com/FlameTN7/dsh-tui-browser-use .

About this plugin

When building agents in dsh-tui, a frequent pain point is that agents can only process plain text and cannot truly see web content. dsh-tui-browser-use fills this gap: it drives a real browser via Playwright and registers 21 browser_* atomic tools covering navigation, interaction, DOM snapshots, visual screenshot analysis, and multi-step autonomous task loops.

The plugin natively integrates DeepSeek vision model screenshot understanding and automatically handles long-page scrolling, tiling, and resolution preservation so critical information survives compression. The session layer supports both persistent login profiles and ephemeral isolated sessions with built-in concurrency locks and conflict-degradation fallback. Security policies block file: protocol and cloud-metadata SSRF by default, restrict file writes to the workspace, and auto-mask sensitive parameters.

If you are already using dsh-tui to build agents for real-world web scenarios such as form filling, data extraction, or multi-step navigation, this plugin upgrades your agent from reading text to seeing pages and taking action, serving as the core browser capability extension in the dsh-tui ecosystem.

Use Cases

  • Agents that browse pages, fill forms, and extract structured data
  • Screenshots sent to a vision model for page layout and content analysis
  • Multi-step navigation with persistent login and autonomous task completion

Best For

  • dsh-tui agent developers
  • Engineering teams that need agents to drive real browsers
  • AI engineers building on DeepSeek vision models