"Zsh" matching MCP servers:

Enables local-first HR payroll computation with statutory social insurance, housing fund, and cumulative IIT calculations, adaptable to arbitrary spreadsheet headers via import adapters and company profiles. Supports safe performance formula evaluation and keeps PII on-device.

An MCP server that gives any agent read/write access to office documents and Jupyter notebooks, with stable cell and shape addresses, bounded windows, and safety safeguards.

Enables creating, reading, editing, and verifying PDF, Word, Excel, and PowerPoint documents through MCP tools that return structured results. It preserves source files by default for edits and supports local stdio operation without network requests.

Enables language models to control a real visible browser with a full set of 24 tools for navigation, clicking, form filling, screenshots, and reading page structure, while preserving login states and supporting configurable browser policies.

A stateful todo-list MCP server that allows adding, listing, completing, and deleting tasks with priority levels, and exposes a markdown resource of the current board.

Enables AI agents to inspect, read, create, format, edit, validate and convert DOCX, XLSX, PPTX and PDF documents through standard OOXML manipulation plus a locally installed R7 converter, without emulating mouse or keyboard input. It also optionally bridges to a document the user has open in R7-Office Desktop over loopback WebSocket.

Enables agents to connect to a hosted workspace and access documents, canvases, slides, sheets, projects, and tickets over MCP from any client or terminal.

Exposes HR business tools — attendance, payroll, organization, and recruiting — as native tools that AI models can call. Connections are secured via MCP OAuth 2.0 (discovery, dynamic client registration, authorization code with PKCE, refresh token rotation), static bearer tokens, or anonymous access.

One secure, self-hosted MCP endpoint connecting AI agents to your real accounts (mail, calendar, files, notes), with per-profile separation, explicit permission enforcement, and an append-only audit log.

Enables DeepSeek Harness agents to operate Feishu/Lark through the official OpenAPI MCP subprocess using app credentials, exposing tools for sending IM messages, reading and writing Bitable tables, and managing docs, calendar, and drive files. A web settings panel configures credentials, tests connectivity, and lists discovered tools.