Install command
Provided
Configure MCP servers in the Claude Desktop app by editing claude_desktop_config.json. Grounded walkthrough covering config file locations, the mcpServers JSON structure, the filesystem server, and how to verify and troubleshoot the connection.
Source-backed facts for citing this resource, derived directly from the registry — also available as plain text for AI assistants.
Decision playbook
Signals are comparatively strong, but you should still validate source, privacy posture, and package provenance for your environment.
0
96
—
No baseline selected
No major trust-signal divergence detected in the current selection.
Confirm ownership and provenance before trusting install instructions.
Source link availableRequired
Open the canonical repository and verify ownership.
Source provenance statusRequired
Marked as first-party.
Metadata reviewed
Registry metadata indicates a reviewed listing.
Validate risk disclosures before installation or API wiring.
Safety notes presentRequired
Review the listed safety guidance before running commands.
Privacy notes presentRequired
Review data handling notes before connecting accounts or secrets.
Trust level risk gateRequired
Trust level does not block evaluation.
Check package metadata and artifact integrity signals.
Install payload available
Install or copy payload is available for review.
Package verification flag
Package marked verified.
Checksum metadata
SHA-256 hash is present.
Use compare context to validate trade-offs before adoption.
Compare tray has multiple entries
Add at least one more entry to compare trust differences.
Baseline comparison available
No baseline peer selected yet.
Diverging trust signals identified
No major trust-signal divergence found.
Setup at a glance
Copy-ready — paste the snippet to get started.
Install command
Provided
Config snippet
Provided
Copy snippet
Provided
Prerequisites
10 to clear
Platforms
2 listed
Difficulty
20/100
Adoption plan
Current risk score 0/100. Use staged verification before broader rollout.
Validate source and review signals before any execution.
Confirm source provenanceRequired
Source URL/provenance metadata is present.
Confirm metadata review state
Listing has review metadata.
Verify install payload
Install/config payload exists and can be inspected.
Confirm safety, privacy, and package integrity signals.
Review safety notesRequired
Safety notes are present.
Review privacy notesRequired
Privacy notes are present.
Verify package integrity metadata
Package verification/checksum metadata is available.
Adopt in controlled steps based on the selected plan.
Run in isolated sandbox firstRequired
Use a constrained sandbox and observe behavior across multiple tasks.
Roll out graduallyRequired
Roll out to a small cohort before wider usage.
Set monitoring and fallback
Define rollback path and monitor errors after adoption.
Evidence readiness
Required evidence gates are covered (6/6 signals complete).
Source repository/provenance is listed.
Required in this preset
Review metadata is present.
Required in this preset
Safety notes are present.
Required in this preset
Privacy notes are present.
Optional in this preset
Package integrity metadata is present.
Optional in this preset
Install payload is available.
Required in this preset
Required evidence gates are covered for this preset.
Decision timeline
6/6 steps complete with no blocking gaps for this preset.
triage
Source/provenance metadata is available.
triage
Review metadata is available.
verify
Safety notes are available.
verify
Privacy notes are available.
verify
Package integrity metadata is available.
rollout
Install payload is available.
No required blockers for this timeline preset.
Prerequisite readiness
10 prerequisites to line up before setup. Have accounts and credentials ready first.
Safety & privacy surface
2 safety and 2 privacy notes across 3 risk areas. Review closely: credentials & tokens, network access.
{
"mcpServers": {
"filesystem": {
"command": "npx",
"args": [
"-y",
"@modelcontextprotocol/server-filesystem",
"/Users/username/Desktop",
"/Users/username/Downloads"
]
}
}
}Claude Desktop can connect to local MCP (Model Context Protocol) servers, which are programs that run on your computer and expose tools Claude can call with your approval — reading files, querying a database, searching the web, and more. You enable these servers by editing a single JSON file, claude_desktop_config.json, that tells Claude Desktop which servers to launch and how to run them.
This guide walks through the official setup flow using the Filesystem Server as the worked example: where the config file lives, the exact mcpServers JSON structure, how to start and verify a server, and how to read the logs when something fails. The same pattern applies to any stdio MCP server.
A few facts worth knowing up front:
npx.node --version
If it is missing, install the LTS release from nodejs.org. If npx fails on Windows, confirm npm is installed globally (%APPDATA%\npm should exist).
The configuration is the same file regardless of how many servers you add. Its path differs by operating system:
| Operating system | claude_desktop_config.json location |
|---|---|
| macOS | ~/Library/Application Support/Claude/claude_desktop_config.json |
| Windows | %APPDATA%\Claude\claude_desktop_config.json |
You do not need to find this file by hand. The fastest way to open it is from inside the app, described in the next section, which creates the file if it does not yet exist.
claude_desktop_config.json if it does not exist and opens it for editing.mcpServers structure below (macOS example), substituting your real username and the directories you want to expose:{
"mcpServers": {
"filesystem": {
"command": "npx",
"args": [
"-y",
"@modelcontextprotocol/server-filesystem",
"/Users/username/Desktop",
"/Users/username/Downloads"
]
}
}
}
On Windows, use escaped backslashes in the paths:
{
"mcpServers": {
"filesystem": {
"command": "npx",
"args": [
"-y",
"@modelcontextprotocol/server-filesystem",
"C:\\Users\\username\\Desktop",
"C:\\Users\\username\\Downloads"
]
}
}
}
Each entry under mcpServers is keyed by a friendly name and describes how to launch a local (stdio) server:
"filesystem" — the display name for the server inside Claude Desktop."command": "npx" — the program used to launch the server. npx runs the Node package without a separate global install."-y" — auto-confirms installation of the server package."@modelcontextprotocol/server-filesystem" — the package name of the server.To pass secrets such as API keys, add an env object to the server entry. The Windows ${APPDATA} workaround from the official troubleshooting guide is a real example of the env shape:
{
"mcpServers": {
"brave-search": {
"command": "npx",
"args": ["-y", "@modelcontextprotocol/server-brave-search"],
"env": {
"APPDATA": "C:\\Users\\user\\AppData\\Roaming\\",
"BRAVE_API_KEY": "your-key-here"
}
}
}
}
Use a directory you are comfortable letting Claude read and modify: the server runs with your user account's permissions, so it can do anything you can do manually with those files.
Once the Filesystem Server is connected, you can ask Claude things like "What files are in my Downloads folder?" or "Save this text to a file on my Desktop." Before any file operation runs, Claude requests your approval, so review each request and deny anything you are not comfortable with.
The config above uses the stdio transport — a local process Claude talks to over standard input/output. Claude Code (the CLI, a separate product) also supports remote transports, which is useful context when choosing how a server connects:
| Transport | Where it runs | Notes |
|---|---|---|
| stdio | Local process on your machine | Ideal for tools needing direct system access or custom scripts; used by claude_desktop_config.json servers. Local processes are not auto-reconnected. |
HTTP (streamable-http) |
Remote service | Recommended for cloud servers; supports OAuth. In Claude Code: claude mcp add --transport http <name> <url>. |
| SSE | Remote service | Server-Sent Events; deprecated in favor of HTTP where available. |
For Claude Desktop, stdio via claude_desktop_config.json is the documented local-setup path covered by this guide.
Server / tools icon not showing up. Restart Claude Desktop completely. Check claude_desktop_config.json for valid JSON. Make sure every path in the config is absolute, not relative. Then read the logs (below). You can also run the server manually to surface errors:
npx -y @modelcontextprotocol/server-filesystem /Users/username/Desktop /Users/username/Downloads
Read the logs. MCP logging is written to:
~/Library/Logs/Claude%APPDATA%\Claude\logsmcp.log holds general connection logging; mcp-server-SERVERNAME.log holds a specific server's stderr. To follow recent logs on macOS/Linux:
tail -n 20 -f ~/Library/Logs/Claude/mcp*.log
Tool calls failing silently. Check Claude's logs for errors, verify the server runs without errors on its own, and restart Claude Desktop.
Windows ${APPDATA} ENOENT error. If a server's log shows an error referencing ${APPDATA} in a path, add the expanded %APPDATA% value to that server's env block (see the brave-search example above), then relaunch Claude Desktop. Confirm npm is installed globally if npx keeps failing.
Claude Desktop MCP Setup side by side with 3 alternatives on trust, install, platform support, and disclosed safety notes — all from reviewed registry metadata.
2 trust signals differ across this comparison (Package trust, Submitter).
Next steps differ across entries — use the actions in the table below to copy install commands and source links per resource.
| Field | Configure MCP servers in the Claude Desktop app by editing claude_desktop_config.json. Grounded walkthrough covering config file locations, the mcpServers JSON structure, the filesystem server, and how to verify and troubleshoot the connection. Open dossier | Master MCP server development from scratch. Create custom Claude Desktop integrations with TypeScript/Python in 60 minutes using production-ready patterns. Open dossier | A practical pre-installation review workflow for Model Context Protocol servers. Inventory tools, resources, prompts, credentials, network reach, storage, and runtime permissions before connecting a server to Claude or another MCP client. Open dossier | Connect existing MCP servers to Claude Code with claude mcp add: stdio, HTTP, and SSE transports, local/project/user scopes, .mcp.json, and /mcp OAuth. Open dossier |
|---|---|---|---|---|
| Next stepsDiffers | ||||
| Trust | ||||
| Review status | ReviewedMaintainer reviewed | ReviewedMaintainer reviewed | ReviewedMaintainer reviewed | ReviewedMaintainer reviewed |
| Package trustDiffers | Package verified | Package not verified | Package not verified | Package not verified |
| Source provenance | Source-backed | Source-backed | Source-backed | Source-backed |
| SubmitterDiffers | — | — | MkDev11 | — |
| Install risk | Low risk | Review first | Review first | Review first |
| Notes | Safety ✓ Privacy ✓ | Safety ✓ Privacy ✓ | Safety ✓ Privacy ✓ | Safety ✓ Privacy ✓ |
| Brand | — | — | — | |
| Category | mcp | guides | guides | guides |
| Source | first-party | source-backed | source-backed | source-backed |
| Author | JSONbored | JSONbored | MkDev11 | JSONbored |
| Added | 2025-10-27 | 2025-10-27 | 2026-06-04 | 2025-10-27 |
| Platforms | Claude CodeClaude Desktop | Claude Code | Claude Code | Claude Code |
| Source repo | — | — | — | — |
| Safety notes | ✓Local MCP servers run as processes on your machine with your user account's privileges, so they can perform any file or system operation you can. Only add servers you trust, and restrict filesystem server paths to the minimum directories the workflow needs. Each Claude Desktop tool call (file write, delete, move) requires your explicit approval before it executes; review every request before approving. | ✓Building and connecting an MCP server runs a local process (or connects to a remote one) that executes tools with your user privileges; only connect servers you trust and review the command and URL first. | ✓An MCP server gives Claude Code model-callable tools that can run on your behalf; a stdio server runs as a local process with your environment access, so treat installing one as running third-party code. Anthropic reviews connectors against its listing criteria before adding them to the Anthropic Directory but does not security-audit or manage any MCP server, so trust verification is your responsibility (per code.claude.com/docs/en/security). Servers that fetch external content can expose you to prompt injection; Claude Code requires trust verification for new MCP servers and prompts for approval before using project-scoped servers from .mcp.json. Prefer least-privilege credentials, scoped directories, and explicit approval for side-effect tools. | ✓Local stdio servers run their command with your user privileges, and remote servers expose model-controlled tools, so connect only servers you trust and review the command, URL, and headers first. |
| Privacy notes | ✓Configured servers receive the prompts, file contents, and directory data needed to run their tools, and may store credentials (API keys, tokens) passed through the env block. The claude_desktop_config.json file holds server commands, file paths, and environment-variable names; keep it private and do not commit it to public repositories. | ✓Connecting servers can pass secrets via --env and OAuth tokens stored in Claude Code's local config; the server process can access whatever data and credentials you grant it. | ✓MCP servers can expose local files, resources, prompt templates, tool arguments, tool outputs, logs, and retrieved data to the connected client and model workflow. For HTTP/SSE transports, OAuth access tokens are stored in the system keychain (macOS) or a credentials file; use "Clear authentication" in the /mcp menu to revoke access. For stdio servers, credentials come from the environment you pass in. Restrict OAuth scopes with oauth.scopes in .mcp.json so a remote server only receives the access it needs, and revoke credentials and delete persisted state when uninstalling a server that had access to private repositories, databases, or files. | ✓Server URLs, command paths, header names, and tool results can expose internal hosts and private data; keep API keys and tokens out of shared `.mcp.json` and use environment-variable expansion or `/mcp` OAuth instead. |
| Prerequisites |
| — none listed |
|
|
| Install | | — | — | — |
| Config | | — | — | — |
| Citations | ||||
| Claim | Unclaimed | Unclaimed | Unclaimed | Unclaimed |
Source-backed guides for putting this to work.
Loading live community signals…
A short, calm digest of reviewed Claude resources. Unsubscribe any time.