dsh-browser-ops
Run the following command in DeepSeek Harness:
dsh plugin install lirui1024k/dsh-browser-ops
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install lirui1024k/dsh-browser-ops in the DeepSeek Harness terminal to install; source code at https://github.com/lirui1024k/dsh-browser-ops. Restart DSH Desktop after installation to activate the browser tools in a new session.
About this plugin
DSH agents often need to open internal systems to check orders, fill and submit forms, or diagnose console and network errors on a page, but until now there was no smooth browser-control pipeline, and hand-crafting test data was repetitive and time-consuming. dsh-browser-ops turns a real browser into the agent's hands and fills that gap.
The plugin ships 27 browser_* tools covering launch, page navigation, snapshots, clicks, form filling, text and table reading, JS evaluation, screenshots, and network diagnostics. Two modes are supported: managed mode spins up an isolated Chrome with a persistent profile, ideal for batch data creation; connect mode attaches to your everyday Chrome to preserve existing sessions. On top of that, the record-and-replay workflow lets you manually click through an operation flow once, save it as a JSON flow, mark input steps as variable slots, then replay with parameters and verify results, eliminating repetitive manual work.
Best suited for developers and QA teams who want their DSH agents to operate web systems, bulk-generate test data, or debug front-end page issues. Recommended on macOS with DSH Desktop.
Use Cases
- Let the DSH agent open internal systems, check orders and submit forms automatically
- Record a manual flow once, then replay with parameters to bulk-generate test data
- Diagnose page console errors and network request failures
Best For
- Developers who need DSH agents to operate real web systems
- QA teams wanting to cut manual test-data creation time
- Engineers automating front-end debugging and issue triage
Related Plugins
A method pack that makes AI coding agents plan against your real baseline, prove completion with fresh evidence, and reduce reworks and unsafe changes.
Turns the DeepSeek Harness session into a captain that builds a durable sub-agent team, splits goals into dependency-aware tasks, and coordinates work via direct messages and a live Web UI.
Gives coding agents design judgment, letting Claude Code, Cursor, and 70+ agents generate and iterate high-quality UI, presentations, and graphics right from the terminal.
Run the Pi ecosystem's plugins on DeepSeek Harness, unmodified, via a compatibility layer that implements Pi's public extension ABI on DSH's native services.