Desktop client for Qwen AI with MCP (Model Context Protocol) support. Manages MCP server connections via Node.js bridge and Rust backend.
This MCP server has critical definition gaps. Tool descriptions are present but minimal (10-50 chars, well below the 194-char baseline). Input schemas are visible but lack depth: most parameters have type and description, but several lack constraints (enums, ranges, patterns). No output schemas are documented. Error handling is absent, no recovery guidance, no categorization. The server bridges Tauri/Electron to MCP but exposes low-level UI operations (dialogs, file pickers, window creation) without clear composition patterns or safety guardrails. Tool names are action-oriented (good), but descriptions do not explain WHEN to use each tool or what happens on failure.
Create a new webview window with a unique label, loading chat.qwen.ai
Retrieve a setting value from the application settings file
Request file access from the user via a file picker dialog
Store a setting value in the application settings file
Display a native dialog box with a title, message, and optional buttons
Signal that a webview has loaded and emit any pending events
No output schemas documented. LLMs cannot plan downstream calls or extract return values. All 6 tools lack documented return types.
Descriptions are too brief (10-50 chars vs. 194-char baseline). They state WHAT but not WHEN to use or what happens on error. E.g., 'Display a native dialog box' lacks context on button handling, return values, or failure modes.
No error handling guidance. Tools do not document what errors are possible, whether they are retryable, or what the LLM should do next. E.g., request_file_access may fail if user cancels, no recovery hint provided.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 46 | <=2025-11-25 | v2 |
Parameter constraints missing. 'buttons' in show_native_dialog is an array of strings with no length limit, format, or validation. 'default_id' is a number with no range (0-N?). LLMs will pass invalid values.
Tool composition unclear. webview_loaded signals completion but does not explain what 'pending events' are or how it chains with other tools. create_new_window loads chat.qwen.ai but does not document what the LLM should do next.