Skip to main content
Glama
matthewdcage

PBS MCP AI Enabled API Server

by matthewdcage

Pharmaceutical Benefits Scheme (PBS) MCP AI Enabled API Server MCP Server

A standalone Model Context Protocol (MCP) server for accessing the Australian Pharmaceutical Benefits Scheme (PBS) API.

About the Author

This PBS MCP server was developed by [Matthew Cage], Founder of https://ai-advantage.au, specialist in Automation, AI Engineering and AI integration and healthcare data systems.

Collaborate with me: https://www.linkedin.com/in/digitalmarketingstrategyexpert/

Related MCP server: OpenFDA

Overview

This project provides a standalone MCP server that allows AI models to access the Australian Pharmaceutical Benefits Scheme (PBS) API, which contains information about medicines, pricing, and availability in Australia.

The project is built for the Public API, but can easily be adapted to the private API if you have been granted developer access.

The PBS API provides programmatic access to PBS data, including medicine listings, pricing, and availability. This MCP server makes it easy to integrate PBS data into AI workflows.

The MCP is available via HTTP and CLI.

Please be aware of the rate limits for the PBS and adjust your request frequency. I recommend a periodic call to store the information you require from the API and update it on a weekly basis.

MCP Server Features MCP Server

This MCP server implements the following Model Context Protocol features:

  • Tools: Provides tools for querying the PBS API endpoints, allowing AI models to access pharmaceutical data

  • Transport Layers: Supports both stdio and HTTP/SSE transport layers

  • Error Handling: Comprehensive error handling for API rate limits and authentication issues

  • LLM Integration: Receives tool calls and prompts directly from LLM components, enabling seamless AI interaction with PBS data

How It Works

The MCP Client MCP Client:

  1. Receives Tool Calls: When an LLM (like Claude) needs pharmaceutical data, it sends a tool call to this server

  2. Processes Prompts: Interprets natural language prompts about medication information

  3. Executes API Queries: Translates the requests into appropriate PBS API calls

  4. Returns Structured Data: Sends back formatted pharmaceutical data that the LLM can use in its responses

This enables AI assistants to access up-to-date PBS information without needing to have this data in their training.

Installation

  1. Clone this repository:

    git clone <repository-url>
    cd pbs-mcp-standalone
  2. Install dependencies:

    npm install
  3. Build the project:

    npm run build

Usage

Starting the Server MCP Server

The PBS MCP server can be run in different modes:

Stdio Mode (Default)

This mode is compatible with the MCP protocol and communicates via standard input/output streams:

npm start

Or use the provided start script:

./start.sh

HTTP Mode with SSE Support

This mode starts an HTTP server with Server-Sent Events (SSE) support:

npm run start:http

Or use the provided start script:

./start.sh http 3000

Where 3000 is the port number to listen on.

Command-Line Interface

The PBS MCP server can also be used as a command-line tool:

npm run cli -- <command>

Or use the provided start script:

./start.sh cli <command>

For example:

./start.sh cli info

Using as a Command-Line Tool

To use this MCP server as a command-line tool:

  1. Build the project:

    npm run build
  2. Run the CLI with the desired command:

    npm run cli -- <command>

    Or use the start script:

    ./start.sh cli <command>

Integrating with MCP Clients MCP Client

This server can be integrated with any MCP-compatible client, such as:

  • Local AI Editors and AI/LLM Servers

  • Other AI assistants that support the Model Context Protocol

  • Custom applications using the MCP client libraries

Client Configuration Example

Here's an example of how to configure this server with an MCP client:

{
  "mcpServers": {
    "pbs-api": {
      "command": "node",
      "args": ["path/to/pbs-mcp-standalone/build/index.js"],
      "env": {
        "PBS_API_SUBSCRIPTION_KEY": "your-subscription-key-here"
      }
    }
  }
}

Accessing the Server from a Client

To access this MCP server from a client:

  1. For Claude Desktop or other MCP-compatible AI assistants:

    • Configure the assistant to use this server as an MCP tool provider

    • The assistant will automatically discover and use the tools provided by this server

    • The LLM can send natural language prompts about medications that will be processed by the server

  2. For custom applications:

    • Use the HTTP API endpoints described below

    • Connect to the SSE endpoint for real-time tool events

    • Or spawn the server process and communicate via stdin/stdout

Example LLM Prompts

The server can interpret various prompts from LLMs, such as:

"Find information about metformin in the PBS"
"What is the PBS code for insulin?"
"List all prescribers who can prescribe antibiotics"
"Get the latest pricing for asthma medications"

These natural language prompts are translated into appropriate PBS API calls.

API Tool Parameters

The PBS API tool can be used with the following parameters:

{
  "endpoint": "prescribers",
  "method": "GET",
  "params": {
    "get_latest_schedule_only": "true",
    "limit": "20"
  }
}

Parameters

  • endpoint (string, required): The specific PBS API endpoint to access (e.g., "prescribers", "item-overview")

  • method (string, optional): HTTP method to use (GET is recommended for most PBS API operations). Default: "GET"

  • params (object, optional): Query parameters to include in the request

  • subscriptionKey (string, optional): Custom subscription key. If not provided, the default public key will be used

  • timeout (number, optional): Request timeout in milliseconds. Default: 30000

HTTP API MCP Server

When running in HTTP mode, the following endpoints are available:

Health Check

GET /health

Returns the status of the server.

List Tools

GET /tools

Returns a list of available tools.

SSE Endpoint

GET /sse

Establishes an SSE connection and sends tool events.

Tool Invocation (SSE)

POST /sse/:toolName

Invokes a tool and sends the result via SSE.

Tool Invocation (REST)

POST /api/:toolName

Invokes a tool and returns the result as JSON.

Command-Line Interface MCP Dev

The PBS MCP server can be used as a command-line tool with the following commands:

List Endpoints

./start.sh cli list-endpoints

Lists all available PBS API endpoints.

Get API Information

./start.sh cli info

Returns information about the PBS API.

Query Prescribers

./start.sh cli prescribers [options]

Options:

  • -l, --limit <number>: Number of results per page (default: 10)

  • -p, --page <number>: Page number (default: 1)

  • -c, --pbs-code <code>: Filter by PBS code

  • -s, --schedule-code <code>: Filter by schedule code

  • -t, --prescriber-type <type>: Filter by prescriber type

  • -f, --fields <fields>: Specific fields to return

  • --latest: Get only the latest schedule

Query Item Overview

./start.sh cli item-overview [options]

Options:

  • -l, --limit <number>: Number of results per page (default: 10)

  • -p, --page <number>: Page number (default: 1)

  • -s, --schedule-code <code>: Filter by schedule code

  • -f, --fields <fields>: Specific fields to return

  • --latest: Get only the latest schedule

Query Any Endpoint

./start.sh cli query <endpoint> [options]

Options:

  • -m, --method <method>: HTTP method (default: GET)

  • -p, --params <json>: Query parameters as JSON string

  • -k, --subscription-key <key>: Custom subscription key

  • -t, --timeout <milliseconds>: Request timeout in milliseconds

Start HTTP Server

./start.sh cli serve [options]

Options:

  • -p, --port <number>: Port to listen on (default: 3000)

Available Endpoints

The PBS API provides several endpoints for accessing different types of data:

  • / - Root endpoint, provides API information and changelog

  • /prescribers - Information about prescribers

  • /item-overview - Detailed information about PBS items

  • /items - Basic information about PBS items

  • /schedules - Information about PBS schedules

  • /atc-codes - Anatomical Therapeutic Chemical (ATC) classification codes

  • /organisations - Information about organisations

  • /restrictions - Information about restrictions

  • /parameters - Information about parameters

  • /criteria - Information about criteria

  • /copayments - Information about copayments

  • /fees - Information about fees

  • /markup-bands - Information about markup bands

  • /programs - Information about programs

  • /summary-of-changes - Summary of changes

For a complete list of endpoints, see the PBS API documentation.

Example Usage

Get API Information

{
  "endpoint": ""
}

Get Prescribers

{
  "endpoint": "prescribers",
  "params": {
    "get_latest_schedule_only": "true",
    "limit": "10"
  }
}

Get Item Overview with Latest Schedule

{
  "endpoint": "item-overview",
  "params": {
    "get_latest_schedule_only": "true",
    "limit": "5"
  }
}

Get Prescribers with Specific PBS Code

{
  "endpoint": "prescribers",
  "params": {
    "pbs_code": "10001J",
    "get_latest_schedule_only": "true"
  }
}

Authentication

The tool uses a subscription key for accessing the PBS API. You can obtain your own key by registering on the PBS Developer Portal.

For development purposes, see the .env.example file for configuration details.

Obtaining a PBS API Subscription Key

To obtain your own PBS API subscription key, follow these steps:

  1. Visit the PBS Data API Portal:

  2. Create an Account:

    • Click on "Sign Up" to create a new account

    • Fill in your details and verify your email address

  3. Subscribe to the PBS API:

    • Once logged in, navigate to the "Products" section

    • Select the "PBS Public API v3" product

    • Click "Subscribe" to request access to the API

  4. Retrieve Your Subscription Key:

    • After your subscription is approved, go to your profile

    • Navigate to "Subscriptions" or "API Keys" section

    • Copy your primary or secondary key

  5. Configure Your Environment:

    • Create a .env file based on the .env.example template

    • Replace your-subscription-key-here with your actual subscription key:

      PBS_API_SUBSCRIPTION_KEY=your-actual-subscription-key

Note: The PBS Public API is rate-limited to one request per 20 seconds. This limit is shared among all users of the public API. For higher rate limits or access to embargo data (future schedules), you may need to apply for special access through the PBS Developer Program.

Limitations

  • The PBS Public API is rate-limited to one request per 20 seconds (shared among all users)

  • Only the current schedule and those published in the past 12 months are available via the Public API

  • Some endpoints require specific parameters to be provided

  • The API structure and endpoints may change over time

Additional Resources

License

This project is licensed under the MIT License - see the LICENSE file for details.


MCP Server MCP Client MCP Dev MCP Enabled ❤️

Available Tools

1 tool
pbs_apiC

Access the Australian Pharmaceutical Benefits Scheme (PBS) API to retrieve information about medicines, pricing, and availability.

ParametersJSON Schema
NameRequiredDescriptionDefault
endpointYesThe specific PBS API endpoint to access (e.g., "prescribers", "item-overview")
methodNoHTTP method to use (GET is recommended for most PBS API operations)GET
paramsNoQuery parameters to include in the request (e.g., {"get_latest_schedule_only": "true"})
subscriptionKeyNoCustom subscription key (if not provided, the default public key will be used)
timeoutNoRequest timeout in milliseconds

TDQS

C2.9/5.0
Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, so the description carries the full burden of behavioral disclosure. It mentions retrieving information, implying a read operation, but doesn't cover critical aspects like authentication requirements (hinted at by the subscriptionKey parameter), rate limits, error handling, or response formats. This is inadequate for a 5-parameter API tool.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single, efficient sentence that directly states the tool's purpose without unnecessary details. It's front-loaded and wastes no words, making it highly concise and well-structured.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness2/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool's complexity (5 parameters, no output schema, no annotations), the description is incomplete. It doesn't explain return values, error conditions, or behavioral traits like authentication or rate limits. This leaves significant gaps for an AI agent to use the tool effectively.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the schema fully documents all parameters. The description adds no specific parameter semantics beyond the general API context. Baseline 3 is appropriate when the schema does the heavy lifting.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool's purpose: 'Access the Australian Pharmaceutical Benefits Scheme (PBS) API to retrieve information about medicines, pricing, and availability.' It specifies the verb ('access'), resource ('PBS API'), and scope ('medicines, pricing, and availability'). However, it doesn't differentiate from siblings (none exist), so it's not a perfect 5.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides no guidance on when to use this tool versus alternatives. It lacks context about prerequisites (e.g., API access, authentication needs) or typical use cases. With no siblings, differentiation isn't needed, but general usage context is missing.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

TDQS

B3.1/5.0
Disambiguation5/5

With only one tool, there is no possibility of ambiguity or overlap between tools. The single tool has a clear and distinct purpose, making it impossible for an agent to misselect between non-existent alternatives.

Naming Consistency5/5

A single tool inherently has perfect naming consistency, as there are no other tools to compare it against. The tool name 'pbs_api' follows a straightforward and descriptive pattern without any conflicting conventions.

Tool Count2/5

A single tool is too few for the apparent scope of accessing a comprehensive API for medicines, pricing, and availability. This minimal toolset likely forces multiple functionalities into one tool, which can be confusing and inefficient for agents trying to perform specific operations like searching, filtering, or retrieving detailed information.

Completeness2/5

The tool surface is severely incomplete for the domain of pharmaceutical benefits information. A single tool cannot adequately cover the expected operations such as searching medicines, retrieving pricing details, checking availability, or filtering by criteria, leading to significant gaps that will cause agent failures in complex tasks.

Maintenance

ActivityInactive
ResponsivenessNo issues

Resources

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Connectors

Related MCP Servers

  • A
    license
    B
    quality
    F
    maintenance
    A Model Context Protocol server providing AI assistants with access to healthcare data tools, including FDA drug information, PubMed research, health topics, clinical trials, and medical terminology lookup.
    7
    78
    126
    MIT
  • A
    license
    A
    quality
    D
    maintenance
    A Model Context Protocol (MCP) server for querying drug information from the OpenFDA API. Features Retrieve drug label information by brand name Retrieve drug information by generic (active ingredient) name Get all brand versions of a generic drug Get adverse event (side effect) reports for a drug
    7
    37
    4
    MIT
  • A
    license
    A
    quality
    D
    maintenance
    A Model Context Protocol (MCP) server that provides seamless access to Moodle developer documentation from moodledev.io. It enables AI assistants and other MCP clients to search, browse, and retrieve Moodle development documentation with version-specific support.
    4
    MIT

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/matthewdcage/pbs-mcp-server'

If you have feedback or need assistance with the MCP directory API, please join our Discord server