"author:irmia2026" matching MCP servers:

Exposes the non-messaging half of the OneBot v11 protocol (currently targeting SnowLuma) as MCP tools, so models can query login info, groups, members, messages and credentials, and perform management actions such as kicking, muting, renaming groups or cards, leaving groups, and approving friend or join requests. A read-only mode can restrict it to query tools only.

Enables an MCP host to observe and control Windows desktop applications deterministically: reading UIA control trees, OCR text and pixel diffs, clicking, typing, scrolling and window management, plus verbatim text reading/appending via ValuePattern/TextPattern. Visual judgment is delegated back to the host model's own multimodal context through annotated screenshots, numbered element cards, grid maps and crosshair zoom-ins, so no separate vision model is needed.