# essentials [Health: Active]

**Category:** 🧠 Knowledge & Memory  
**Repository:** https://github.com/fieldcure/fieldcure-mcp-essentials  
**GitHub Stars:** 0  
**Views:** 0  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/essentials

## Description
Essential MCP server: HTTP, web search, downloads, Wolfram|Alpha, shell, JS, file I/O, memory.

## Tools
Capabilities this server exposes over MCP:

- **http_request** — Full HTTP client (GET/POST/PUT/DELETE/PATCH/HEAD) with custom headers and body
- **web_search** — Search the web and return snippets (title, URL, description)
- **web_fetch** — Fetch a URL and extract content as Markdown — HTML pages and documents (PDF, DOCX, HWPX, PPTX, XLSX)
- **download_file** — Download URL content to disk with a configurable download directory, 100 MB limit, and atomic save
- **run_command** — Execute shell commands with working directory, environment variables, shell selection, and output truncation flags
- **run_javascript** — Sandboxed JavaScript execution (Jint) for math, data processing, JSON, regex
- **wolfram_alpha** — Wolfram|Alpha Full Results API — symbolic math, plots, unit conversions, constants; MathML passes through for native rendering
- **get_environment** — System info — local time, timezone, OS, hostname, username, .NET version
- **read_file** — Read files — text with offset/limit, documents (PDF, DOCX, HWPX, PPTX, XLSX) parsed to Markdown
- **write_file** — Write or append text to files with auto directory creation
- **search_files** — Search files by glob pattern and content (grep-like)
- **remember** — Store a key-value memory (persisted in SQLite)
- **forget** — Delete memories by key or keyword search
- **list_memories** — Search and list stored memories with FTS5 and pagination
- **search_news** — Search recent news articles via Google News
- **search_images** — Search images with size/type filtering
- **search_scholar** — Search academic papers with citation counts
- **search_patents** — Search patent documents with inventor/assignee filtering
- **set_search_engine** — Switch the active engine (`bing`, `duckduckgo`, `serper`, `tavily`, `serpapi`) at runtime. Paid-engine API keys resolve lazily via env var or MCP Elicitation on the next search; the switch itself takes only the engine name. Emits `notifications/tools/list_changed` on success.
- **get_search_engine** — Return the currently active engine and its category capabilities. Read-only; use this to reflect live engine state in host UIs or to check category support before calling a category search tool.

## Claude Desktop Quick Installation
No confirmed install details yet, so no config is published. Follow the setup steps in the repository above.

## Documentation

## What essentials MCP server does

The essentials MCP server provides a broad collection of tools for research, computation, local automation, and persistent context. Its capabilities include raw HTTP requests, web search, page and document retrieval, file downloads, shell commands, sandboxed JavaScript, Wolfram|Alpha calculations, system information, file operations, and memory stored in SQLite.

Search tools cover general web results as well as recent news, images, academic papers, and patents. Category searches depend on the active search engine. The server can switch between Bing, DuckDuckGo, Serper, Tavily, and SerpApi at runtime, then report the selected engine and its supported categories.

## How it works

The essentials MCP server uses the standard stdio transport. An MCP client starts the `fieldcure-mcp-essentials` process and communicates with it using JSON-RPC over standard input and output. The project is built with C# and the official MCP C# SDK.

Web retrieval tools serve different purposes. `http_request` returns raw HTTP responses for API calls and other requests. `web_search` returns result snippets, while `web_fetch` converts HTML pages and supported documents into Markdown. `download_file` preserves the original bytes on disk and supports a 100 MB limit.

`web_fetch` and `read_file` can parse PDF, DOCX, HWPX, PPTX, and XLSX files into Markdown. PDF extraction reads the text layer only, so scanned PDFs without embedded text do not produce OCR content. HTTP requests, web fetches, and downloads block private IP ranges and loopback addresses.

## Setup and configuration

Install the global .NET tool with:

```bash
dotnet tool install -g FieldCure.Mcp.Essentials
```

The server requires .NET 8.0 or later. After installation, the `fieldcure-mcp-essentials` command is available to configure in an MCP client. Building from source requires cloning the repository and running `dotnet build`.

Bing is the default search engine and does not require an API key. The server can also use Serper, SerpApi, or Tavily. Set `ESSENTIALS_SEARCH_ENGINE` and `ESSENTIALS_SEARCH_API_KEY`, pass the equivalent command-line options, or let the server detect keys from `SERPER_API_KEY`, `SERPAPI_API_KEY`, and `TAVILY_API_KEY`. Paid-engine credentials are resolved when needed for a search.

## Tools and capabilities

The essentials MCP server exposes tools for:

- Sending GET, POST, PUT, DELETE, PATCH, and HEAD requests with headers and bodies.
- Searching the web, news, images, scholarly papers, and patents.
- Fetching web content, downloading files, and parsing supported office documents.
- Running shell commands with a working directory, environment variables, shell choice, and output limits.
- Executing JavaScript in Jint with timeout, statement-count, and recursion limits.
- Querying Wolfram|Alpha for symbolic mathematics, plots, unit conversions, and constants.
- Reading, writing, appending, and searching local files.
- Inspecting local time, timezone, OS, hostname, username, and .NET version.
- Storing, deleting, and searching persistent key-value memories with SQLite and FTS5.

## Limitations and notes

The essentials MCP server includes tools that can modify the local system, including shell execution, file writing, downloads, and memory deletion. Review tool arguments and client permissions before allowing agent use.

Relative download paths resolve beneath the configured download directory, which defaults to `~/Downloads/mcp`; absolute paths are allowed except for protected system directories. Downloads use a temporary file followed by an atomic move, reducing the chance of leaving a partial destination file.

Category search tools remain registered even when the selected engine does not support them, but invocation returns an error explaining that another engine must be selected. Serper, SerpApi, and Tavily support different category sets, and Bing and DuckDuckGo do not provide the listed category searches. Wolfram|Alpha is optional and requires its configured service access.

_Full upstream README: https://allmcps.com/mcp/essentials/readme_