mst-mcp
by vipulsutar11
README.md
# MST-MCP Server
An official Model Context Protocol (MCP) server for the **MST Chain** ecosystem. This server provides tools for LLMs (like Claude) to search, list, and retrieve MST developer documentation (APIs, wallets, transactions, authentication, etc.).
## 🚀 Deployed Endpoint
* **Base URL (SSE Transport):** `http://<YOUR_AWS_EC2_IP_OR_DOMAIN>:<PORT>/sse`
* **Favicon / Branding:** `http://<YOUR_AWS_EC2_IP_OR_DOMAIN>:<PORT>/favicon.png`
---
## 🛠 Exposed Tools
All tools are read-only lookup tools with custom annotations:
1. **`list_documents`**
* **Description:** Lists all available developer documentation files in the server.
* **Annotations:** `readOnlyHint: true`, `title: "List Documents"`
2. **`read_document`**
* **Description:** Reads and returns the contents of a specific documentation file (e.g., `SDK.txt`).
* **Annotations:** `readOnlyHint: true`, `title: "Read Document"`
3. **`search_documents`**
* **Description:** Searches for a keyword or query across all files and returns matching lines with line numbers.
* **Annotations:** `readOnlyHint: true`, `title: "Search Documents"`
---
## 💡 Example Prompts to Try in Claude
After connecting this server as a connector, you can ask Claude:
* **Example 1 (Listing docs):**
> "Check my mcp is in working and list all documentation files."
* **Example 2 (Searching keywords):**
> "Search the developer docs for wallet integration details."
* **Example 3 (Retrieving document content):**
> "Read the documentation on transaction structure."
---
## 🏗 Setup & Deployment
### Local Development
To run the server locally:
1. Initialize virtual environment and install dependencies:
```bash
pip install -r requirements.txt
```
2. Start the server:
```bash
python server.py
```
### Deploying on AWS EC2
We have provided a dedicated deployment guide for AWS EC2.
Please refer to the [`DEPLOYMENT.md`](./DEPLOYMENT.md) file for comprehensive, step-by-step instructions on setting up your EC2 instance natively or via Docker.
TDQS
A3.8/5.0
Scored across 3 tools
Disambiguation5/5
Each tool targets a distinct operation: listing filenames, reading a specific file, and searching content. There is no overlap in functionality.
Naming Consistency5/5
All tools follow a consistent verb_noun pattern (list_documents, read_document, search_documents), making the naming predictable and easy to understand.
Tool Count4/5
With only 3 tools, the server is minimal but well-scoped for its purpose of documentation access. While it could potentially include a tool for metadata, the current count is reasonable.
Completeness4/5
The server covers the core operations of listing, reading, and searching documentation files. Minor gaps like document metadata retrieval are not essential for basic usage, so the surface feels complete.
Maintenance
ActivityMaintained
ResponsivenessNo issues