Skip to main content
Glama
Zilliqa Insights MCP Server

Zilliqa Insights MCP Server

Official
by Zilliqa

Zilliqa Insights MCP Server

The Zilliqa Insights Model Context Protocol (MCP) Server is a component designed to provide LLM interaction with Zilliqa validator nodes observability metrics.

Prerequisites

Related MCP server: THORChain MCP Server

Usage

You can ask your MCP client natural language questions about the Zilliqa validator nodes. Here are a few examples:

  • "What is the zil address for Huobi?"

  • "What is the zil stake for Binance in October 2025?"

  • "How reliable was Zillet at proposing blocks last week?"

  • "What were the total ZIL rewards for Moonlet yesterday?"

  • give me the top 10 validators with more stake

  • give me the top 10 validators with more rewards

Your MCP client will translate these questions into the appropriate tool calls to fetch the data from the MCP server.

Available Tools

This server exposes several tools that query validator performance and status metrics. These tools act as a proxy, connecting to the downstream observability-mcp server, which is part of the gcloud-mcp, to retrieve data from Google Cloud Monitoring.

  • listValidators()

  • Lists all known validators and their metadata: name, public_key, address, and zil_address.

  • getTotalValidatorEarnings(validator, startTime?, endTime?)

    • Retrieves the total ZIL rewards earned by a specific validator within a given time frame. Defaults to the last hour if no time is specified.

  • getValidatorEarningsBreakdown(validator, startTime?, endTime?)

    • Provides a detailed breakdown of a validator's earnings, separating rewards from block proposals and cosignatures.

  • getValidatorStake(public_key)

    • Fetches the total amount of ZIL currently delegated (staked) to a validator, which represents their weight in the consensus mechanism.

  • getProposerSuccessRate(public_key, startTime?, endTime?)

    • Calculates the success rate for a validator when they are tasked with proposing a new block. This is a critical indicator of node stability and network latency.

  • getCosignerSuccessRate(public_key, startTime?, endTime?)

    • Measures the validator's success rate for cosigning (attesting to) blocks proposed by others. This demonstrates consistent uptime and connectivity.

  • getTopValidatorsByEarnings(startTime?, endTime?, limit?)

  • Retrieves the top N validators ranked by total ZIL earnings within a given time frame. Defaults to the last hour and top 5 when not specified.

  • getTopValidatorsByStake(startTime?, endTime?, limit?)

  • Retrieves the top N validators ranked by current delegated stake (GAUGE metric, latest value per validator). Defaults to the last hour and top 5.

  • getTopProposerSuccessRate(startTime?, endTime?, limit?)

  • Retrieves the top N validators ranked by proposer success rate over a time frame (successful proposals divided by total proposals). Defaults to the last hour and top 5.

  • getTopCosignerSuccessRate(startTime?, endTime?, limit?)

  • Retrieves the top N validators ranked by cosigner success rate over a time frame (successful cosignatures divided by total cosignatures). Defaults to the last hour and top 5.

Development

Follow these instructions to get the server running on your local machine for development and testing purposes.

1. Installation

First, install the project dependencies using npm:

npm install

2. Build

Next, compile the TypeScript source code into JavaScript:

npm run build

This will create a build directory containing the distributable files.

3. Running the Server

The server can operate in two modes:

  • stdio mode: The server communicates over standard input/output. This is typically used for direct interaction with the MCP server on the same machine. This is the default option.

  • http mode: The server exposes an HTTP API, allowing for remote communication and management from other services.

node build/index.js

To run the server in HTTP streamable mode add the --http flag:

node build/index.js --http

4. Configuring the LLM settings

Add this configuration in the LLM local settings to test the MCP server. This is an example for Gemini:

"mcpServers": {
  "insights-local": {
    "httpUrl": "http://localhost:3001/mcp"
  }
}

Note: The Zilliqa Insights MCP server connects to an instance of the gcloud MCP server referred in the OBSERVABILITY_MCP_URL variable to retrieve the validators data. These observability metrics are restricted and not publicly available.

Deployment

Kubernetes

The Kubernetes manifests for deploying this server are located in the cd/ directory. Environment-specific configurations can be found in cd/overlays/.

Production environment

A production version of this server is automatically deployed via GitHub Actions pipelines. The deployment is triggered on the creation of a new release and it is accessible at the following URL:

Configuring the LLM settings

Add this configuration in the LLM local settings to test the MCP server. This is an example for Gemini:

"mcpServers": {
  "insights": {
    "httpUrl": "https://insights.mcp.zilliqa.com/mcp"
  }
}

Available Tools

11 tools
get_cosigner_success_rateA

Measures a validator's performance when tasked with cosigning (attesting to) a block proposed by another validator. Cosigning is the most frequent duty. A high success rate demonstrates consistent uptime and connectivity. Even a small dip here can lead to a noticeable reduction in rewards over time. If startTime and endTime are not provided, it defaults to the last hour.

Inputs: validator (required), startTime, endTime.

get_proposer_success_rateA

Measures a validator's performance specifically when tasked with proposing a new block. A 100% proposer success rate is the gold standard. A missed proposal means a delay in the chain and lost rewards for that validator. This metric is a critical indicator of a validator's node stability and network latency. If startTime and endTime are not provided, it defaults to the last hour.

Inputs: validator (required), startTime, endTime.

get_top_cosigner_success_rateB

Gets the top N validators with the highest cosigner success rate within a time frame. Defaults to the last hour and top 5.

Inputs: startTime, endTime, limit.

get_top_proposer_success_rateA

Gets the top N validators with the highest proposer success rate within a time frame. Defaults to the last hour and top 5.

Inputs: startTime, endTime, limit.

get_top_validators_by_earningsA

Gets the top N validators with the highest total earnings within a time frame. Defaults to the last hour and top 5.

Inputs: startTime, endTime, limit.

get_top_validators_by_stakeB

Gets the top N validators with the highest current stake within a time frame. Defaults to the last hour and top 5.

Inputs: startTime, endTime, limit.

get_total_validator_earningsB

Gets the total earnings for a specific validator. If startTime and endTime are not provided, it defaults to the last hour.

Inputs: validator (required), startTime, endTime.

get_validator_earnings_breakdownA

Provides a detailed breakdown of a validator's earnings, separating rewards from block proposals and cosignatures. If startTime and endTime are not provided, it defaults to the last hour.

Inputs: validator (required), startTime, endTime.

get_validator_infoA

Gets all available information for a validator (name, public key, address, zil_address) by providing any one of those identifiers.

Inputs: validator (required).

get_validator_stakeB

Retrieves the total amount of ZIL currently delegated to a validator, representing their weight in the consensus mechanism.

Inputs: validator (required).

list_validatorsA

Lists all known validators with their name, public_key, address, and zil_address.

Inputs: no named parameters.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.9/5.0

    Scored across 11 tools

    Disambiguation5/5

    Each tool targets a distinct validator metric or listing function. There is no overlap between cosigner/proposer success rates, top lists, earnings, stake, and information lookups.

    Naming Consistency5/5

    All tools use 'get_' prefix except 'list_validators', which is a common exception. The pattern is uniform and predictable, aiding agent selection.

    Tool Count5/5

    11 tools cover the core validator insights domain adequately without being excessive. Each tool serves a clear purpose within the scope of performance and earnings monitoring.

    Completeness4/5

    The set covers individual lookups, top rankings, and earnings breakdowns. Minor gaps exist (e.g., no tool for validator commission rates or historical performance trends), but core workflows are supported.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • F
      license
      B
      quality
      Not graded
      maintenance
      Enables AI-powered blockchain data queries and analysis through the Native Indexer (NIX) system. Supports querying blocks, transactions, account information, and network status across various blockchain networks.
      3
      -
    • F
      license
      Not graded
      quality
      D
      maintenance
      Provides tools for querying onchain data across 12+ blockchain networks, including token balances, transaction analysis, and smart contract security auditing. It enables users to interact with multiple EVM-compatible chains and perform deep contract evaluations through natural language interfaces.
      1
      -
    • A
      license
      A
      quality
      B
      maintenance
      Enables AI assistants to access real-time on-chain crypto analytics, whale tracking, and market metrics through natural language queries. It provides access to over 245 endpoints for comprehensive data analysis of assets like Bitcoin, Ethereum, and stablecoins.
      7
      85 npm
      9
      MIT