MCP Prompts Server

How to use

You connect to an MCP server from your client application or IDE by configuring an MCP server entry with a command or URL, then invoke tools to perform actions such as reading files, storing context, or uploading data to AWS. Start by choosing a deployment method that fits your workflow: run a local stdio MCP server during development or host a containerized server for production. Use the provided tool calls to interact with the server, and monitor results through your client’s feedback loop.

How to install

Prerequisites you need before installing: Node.js 18+ or a compatible runtime, Docker, AWS CLI for AWS integrations, and Git.

Step by step installation and setup you can follow locally:

# Prerequisites
# Ensure Docker, Node.js 18+, AWS CLI, and Git are installed on your system.

# Step 1: Clone the project
git clone https://github.com/sparesparrow/mcp-prompts.git
cd mcp-prompts

# Step 2: Build the AWS MCP server image (example for a local container)
docker build -f Dockerfile.aws -t mcp-aws-server .

# Step 3: Prepare environment for running the AWS MCP server
cp .env.example .env
# Edit .env to set MCP_SERVER_PORT and any AWS-related variables (e.g., AWS_REGION)

# Step 4: Run the MCP server
docker run -p 3000:3000 --env-file .env mcp-aws-server

# Optional: If you prefer to use a specific configuration for the AWS server, adjust env vars accordingly

Configuration and usage notes

You will typically configure your MCP client with a server entry that points to a running container or to an HTTP URL if the server is hosted remotely. The AWS MCP server example uses an environment file to set up ports and region information, and the command shown runs the container with that environment.

Environment variables commonly used in this setup include MCP_SERVER_PORT to specify the port the server listens on, and AWS_REGION to select the AWS region for services like S3, Lambda, and DynamoDB. Ensure these are present in the environment file you supply to the container.

Examples

Example interactions you can perform via the MCP client include reading a file, storing conversation context in memory, and uploading data to S3 through the AWS MCP server.

Code snippets below illustrate how you might call common tools once your client is configured to talk to the MCP server.

Notes on deployment options

You can deploy MCP servers as Docker containers for quick development or in a more robust container orchestration environment for production. The AWS integration server supports typical AWS services via MCP endpoints, enabling you to perform object storage, serverless function invocations, and NoSQL data operations from your prompts.

Troubleshooting

If the MCP server fails to start, verify that required environment variables are set and Docker is running. Check the container logs for errors and ensure the port you configured is not in use.

add_prompt

Create a new prompt with metadata, including name, content, and optional tags and variables.

get_prompt

Retrieve a prompt by its unique identifier.

list_prompts

List prompts with optional filtering by tags or search terms.

update_prompt

Update fields on an existing prompt by ID.

delete_prompt

Remove a prompt by its ID.

apply_template

Apply variable values to a prompt template to generate a concrete prompt.

get_stats

Return statistics about stored prompts including counts and available tags/categories.