dsh-erp
Run the following command in DeepSeek Harness:
dsh plugin install GuoMonth/dsh-erp
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install GuoMonth/dsh-erp in DeepSeek Harness to install this plugin; source code at https://github.com/GuoMonth/dsh-erp
About this plugin
The menu structure, field semantics, and business relationships of an ERP often exist only in an operator's head and inside a single browser session, so every AI conversation begins from scratch. dsh-erp lets DeepSeek Harness observe your ERP through an already-logged-in browser window, recording menu hierarchies, page fields, and business data into a local SQLite database and Markdown exports. Later conversations can reuse that knowledge without reopening a browser or re-exploring the interface.
By design it stops far short of being an unattended transaction engine. Every click, fill, selection, or query triggers a separate approval prompt; observation grants expire after ten minutes; actions rely on short-lived snapshot references; credential fields and hidden inputs are excluded from snapshots by default. The plugin makes no assumption about a specific ERP vendor, API path, or token format; compatibility depends entirely on whether your interface can be read correctly by the Playwright runtime.
Built for operators or small teams who already use DSH and want the AI to gradually learn their particular ERP. You handle login and step-by-step confirmations; the plugin handles recording, cross-referencing, and exporting. All knowledge stays in your local directory, is never bundled into the npm release, and is shared only when you explicitly export it.
Use Cases
- Observe ERP menus and fields via the browser to build operational knowledge
- Reuse learned screen structures and business data in later AI conversations
- Step-by-step approved learning to avoid unattended misclicks
Best For
- Operators or small teams already using DeepSeek Harness
- Users who want AI to gradually learn their specific ERP interface
- Teams needing localized, controllable, and auditable AI-assisted learning
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.