MCP Scaleway Functions

io.github.cyclimsePublished by GitHub account cyclimsev0.3.4Updated Oct 24, 2026

Unofficial MCP Server for Scaleway Serverless Functions

VerifiedSTDIODesktop onlyCloud & Infrastructure Developer Tools

Server details

Identifier: io.github.cyclimse/mcp-scaleway-functions
Transport: STDIO
Runtime: Desktop only
Trust: Verified
Published: Sep 16, 2026

Security note

MCP servers receive tool arguments from a conversation and may perform external actions. Review what a server can do before you enable it, and give it only the access it needs.

The registry verified the GitHub account cyclimse for this listing. That does not show the account is the product's maker.

SourceWeft's checks cover whether the listing is complete and the endpoint answered; they do not review the server's code or what its tools do.

Add a badge to your README

Show readers this MCP server is listed on SourceWeft. The badge links to this page.

[![Listed on SourceWeft](https://sourceweft.com/badge/mcp/io.github.cyclimse%2Fmcp-scaleway-functions.svg)](https://sourceweft.com/mcp/io.github.cyclimse%2Fmcp-scaleway-functions?utm_source=readme&utm_medium=badge)

Overview

AI-generated overview

Unofficial MCP server that lets an assistant manage and deploy Scaleway Serverless Functions: namespaces, functions, code, runtimes and logs.

What you can do: Create, list and delete function namespaces
Create, deploy, update and delete functions, and list available runtimes
Download a function's code and fetch its logs
Add dependencies to a local function, including ones needing Docker
When to use it: Useful when you want an assistant to handle Scaleway Serverless Functions end to end from your editor or coding agent: scaffolding a namespace, deploying or updating code, pulling an existing function down to work on it, or checking logs.
Requirements: Runs locally as a binary or container; the README describes SSE on localhost:8080 by default or stdio via a flag. Needs Scaleway credentials, read by default from the standard Scaleway config file, with further settings such as SCW_DEFAULT_REGION via Scaleway environment variables. Docker is needed for dependencies that rely on native code.

Before you install

Unofficial and not affiliated with Scaleway. It can create, update and delete namespaces and functions, and the README says its safety measures against destructive actions are not foolproof. It uses your Scaleway credentials, and debug logging records all requests and responses to and from the Scaleway API.

Installation

In SourceWeft

Open MCP Scaleway Functions in the dashboard and add it to a workspace.
Enable the server for the chats that should use its tools.

Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.

Other MCP clients

Follow the launch instructions in the repository.

README

MCP Scaleway Functions

Model Context Protocol (MCP) server to manage and deploy Scaleway Serverless Functions using the Model Context Protocol standard.

[!CAUTION] This project is unofficial and not affiliated with or endorsed by Scaleway. Some small safety measures are in place to prevent the LLM from doing destructive actions, but they're not foolproof. Use at your own risk.

Getting Started

Download the latest release from the releases page or build it from source using Go.

Run the MCP server:

By default, the MCP server runs with the SSE transport on http://localhost:8080, but you can also change it to use Standard I/O (stdio) transport via the --transport stdio flag.

Then, configure your IDE or tool of choice to connect to the MCP server. Here are some examples:

VSCode (sse example)

Add a new server configuration in your .vscode/mcp.json file:

json

{	"servers": {		"mcp-scaleway-functions": {			"url": "http://localhost:8080",			"type": "http",		}	},}

{	"servers": {		"mcp-scaleway-functions": {			"url": "http://localhost:8080",			"type": "http",		}	},}

Crush (stdio example)

Crush is an open-source coding agent that supports MCP. You can find more information about in the Crush repository.

Add a new server configuration in your ~/.config/crush/crush.json file:

json

{  "$schema": "https://charm.land/crush.json",  "mcp": {    "scaleway-functions": {      "type": "stdio",      "command": "mcp-scaleway-functions",      "args": ["--transport", "stdio"],      "timeout": 600,      "disabled": false    }  }}

{  "$schema": "https://charm.land/crush.json",  "mcp": {    "scaleway-functions": {      "type": "stdio",      "command": "mcp-scaleway-functions",      "args": ["--transport", "stdio"],      "timeout": 600,      "disabled": false    }  }}

You can even use Crush with Scaleway Generative APIs by adding a new provider in the same ~/.config/crush/crush.json file:

jsonc

{  "mcp": {	// ... see above ...  },  "providers": {    "scaleway": {      "name": "Scaleway",      "base_url": "https://api.scaleway.ai/v1/",      "type": "openai",	  // To fetch from environment variables, use the `$VAR_NAME` syntax.	  // Note: this key requires the "GenerativeApisModelAccess" permission.      "api_key": "$SCW_SECRET_KEY",      "models": [        {          "name": "Qwen coder",          "id": "qwen3-coder-30b-a3b-instruct",          "context_window": 128000,          "default_max_tokens": 8000        }      ]    }  }}

{  "mcp": {	// ... see above ...  },  "providers": {    "scaleway": {      "name": "Scaleway",      "base_url": "https://api.scaleway.ai/v1/",      "type": "openai",	  // To fetch from environment variables, use the `$VAR_NAME` syntax.	  // Note: this key requires the "GenerativeApisModelAccess" permission.      "api_key": "$SCW_SECRET_KEY",      "models": [        {          "name": "Qwen coder",          "id": "qwen3-coder-30b-a3b-instruct",          "context_window": 128000,          "default_max_tokens": 8000        }      ]    }  }}

That's it 🎉! Have fun vibecoding and vibedevoopsing as you please.

Configuration

By default, the MCP server reads from the standard Scaleway configuration file located at ~/.config/scw/config.yaml.

Further configuration can be done via the Scaleway environment variables to configure the MCP server.

For instance, you can set a region to work in via the SCW_DEFAULT_REGION environment variable.

Available Tools

Tool	Description
`create_and_deploy_function_namespace`	Create and deploy a new function namespace.
`list_function_namespaces`	List all function namespaces.
`delete_function_namespace`	Delete a function namespace.
`list_functions`	List all functions in a namespace.
`list_function_runtimes`	List all available function runtimes.
`create_and_deploy_function`	Create and deploy a new function.
`update_function`	Update the code or the configuration of an existing function.
`delete_function`	Delete a function.
`download_function`	Download the code of a function. This is useful to work on an existing function.
`fetch_function_logs`	Fetch the logs of a function.
`add_dependency`	Add a dependency to a local function. Useful for dependencies that rely on native code and therefore need Docker to be installed.

Debugging

You can enable debug logging by using the --debug flag when starting the MCP server. This will log all requests and responses to/from the Scaleway API.

To configure the log level, use the --log-level flag (default is info). Available log levels are: debug, info, warn, error.

Logs are stored in the $XDG_STATE_HOME/mcp-scaleway-functions directory (usually ~/.local/state/mcp-scaleway-functions).

Development

Running tests:

Generating mocks:

Source: README.md at commit c68860d

Tools

SourceWeft has not started this package in its sandbox yet, so its tools are not known.

Version history

v0.3.4LatestSep 16, 2026

Score

50% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

Reached in the last 30 days25 pts
Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.
Auth declared10 pts
Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.
Tools listed15 pts
Answer tools/list without a credential, or ship a package that starts in a sandbox without one.
DescriptionRequired15 pts
Logo10 pts
Documentation15 pts
License10 pts