Provides deployment and hosting infrastructure for the MCP server, enabling it to run on Cloudflare's edge network with potential KV storage integration for tag caching
Supports deployment via Git-based workflows, allowing users to deploy the MCP server directly from a Git repository
Enables direct deployment from GitHub repositories using Cloudflare's 'Deploy from Git' functionality or the deploy button
ASMR-One MCP Server (Cloudflare Worker)
This project implements an MCP (Model Context Protocol) server for fetching ASMR work information from asmr-one.com, designed to run on Cloudflare Workers.
Features
- Search ASMR: Search for ASMR works by keyword, circle, tag, or voice actor.
- Random ASMR: Get a random ASMR work recommendation.
- Returns links in
asmr.one
ordlsite
format.
Deploy to Cloudflare
You can deploy this worker to your Cloudflare account using several methods:
Method 1: Deploy with Cloudflare Button (Simplest)
This button initiates a guided deployment process. It should automatically detect the project type and settings from wrangler.toml
.
Method 2: Manual "Connect to Git" in Cloudflare Dashboard
If you manually set up a "Deploy from Git" connection for this Worker in your Cloudflare dashboard (e.g., via a link like https://dash.cloudflare.com/YOUR_ACCOUNT_ID/workers-and-pages/create/deploy-to-workers?repository=...
):
- Project Name: Choose a name (e.g.,
asmr-one-mcp-worker
). - Production Branch:
main
. - Root Directory:
/
(ifwrangler.toml
is in the root of your repository). - Build command (Optional):
- You can often leave this blank. Cloudflare's build system will use
wrangler.toml
to deploy. - If a command is required, or to ensure all dependencies are explicitly installed in the build environment, use:
npm install
- You can often leave this blank. Cloudflare's build system will use
- Build output directory (Optional): Leave this blank. This is not typically used for Worker deployments in the same way it is for static sites (Cloudflare Pages).
- Cloudflare will then use
wrangler
(based on yourwrangler.toml
) to build and deploy your Worker upon commits to the production branch.
Method 3: Using Wrangler CLI (Manual Deployment)
This method gives you the most control and is useful for local testing and manual deployments.
- Clone the repository:
- Install dependencies:
- Log in to Cloudflare (if you haven't already):
- Deploy the Worker:(This script in
package.json
runswrangler deploy
)
After Deployment: Cloudflare will provide you with a unique URL for your worker (e.g., https://your-worker-name.your-username.workers.dev
). You will need this URL to configure it as a remote MCP server.
Configuration
The primary configuration is done via the wrangler.toml
file.
Tag Map (Tag Caching)
The server fetches and caches ASMR tags to improve search accuracy for aliased or partial tags. Currently, this cache is in-memory per worker instance.
For more robust and persistent caching in a production environment, consider using Cloudflare KV:
- Create a KV Namespace in your Cloudflare dashboard.
- Uncomment and configure the
[[kv_namespaces]]
section inwrangler.toml
: - The code in
src/index.ts
has commented-out lines for KV integration. You would need to uncomment and potentially adapt theloadTagsIfNeeded
function to useenv.ASMR_TAGS_KV
. - Re-deploy your worker after making these changes.
Local Development (using Wrangler CLI)
- Ensure you have cloned the repository and installed dependencies as shown in "Method 3" above.
- Start the local development server:This will typically start a server on
http://localhost:8787
for testing your Worker locally.
Using as an MCP Server
Once deployed, you can add this worker as a remote MCP server in your MCP client's settings (e.g., mcp_settings.json
).
Example configuration:
Make sure the /mcp
path in the URL matches how your worker handles requests (currently, the provided src/index.ts
expects MCP requests on the root path of the worker, so /mcp
might not be needed unless you add routing). If your worker handles MCP requests at the root, the URL would be https://your-worker-name.your-username.workers.dev
. The current src/index.ts
does not implement a specific /mcp
route, it processes POST requests to the worker's root.
Project Structure
src/index.ts
: The main worker code.wrangler.toml
: Cloudflare Worker configuration file.package.json
: Project dependencies and scripts.tsconfig.json
: TypeScript configuration.
This server cannot be installed
A Cloudflare Worker that implements an MCP server for searching and retrieving ASMR work information from asmr-one.com, allowing users to search by keyword, circle, tag, or voice actor and get random recommendations.
Related MCP Servers
- -securityAlicense-qualityA lightweight MCP server for managing DNS records, purging cache, and interacting with the Cloudflare API through natural language commands.Last updated -13TypeScriptMIT License
- -securityAlicense-qualityAn MCP server that allows using natural language to manage Cloudflare resources (Workers, KV, R2, D1) through Claude Desktop, VSCode, and other MCP clients.Last updated -85TypeScriptApache 2.0
- -securityAlicense-qualityAn MCP server that enables web searches using a SearxNG instance, allowing MCP-compatible applications like Goose to perform internet searches.Last updated -PythonGPL 3.0
- -securityFlicense-qualityA deployable MCP (Model Context Protocol) server on Cloudflare Workers that doesn't require authentication, allowing users to create and access custom AI tools through Claude Desktop or the Cloudflare AI Playground.Last updated -TypeScript