Back to Home
manish6007 icon

mcp_servers

by manish6007

Overview

A production-grade Model Context Protocol (MCP) server combining Amazon Redshift SQL query capabilities and a Knowledgebase vector store for AI agents.

Installation

Run Command
mcp-server-http

Environment Variables

  • REDSHIFT_CLUSTER_ID
  • REDSHIFT_DATABASE
  • REDSHIFT_HOST
  • REDSHIFT_RESULTS_BUCKET
  • KNOWLEDGEBASE_S3_BUCKET
  • POSTGRES_SECRET_NAME
  • POSTGRES_HOST
  • POSTGRES_DATABASE
  • POSTGRES_USER
  • POSTGRES_PASSWORD
  • AWS_REGION
  • AWS_ACCESS_KEY_ID
  • AWS_SECRET_ACCESS_KEY
  • AWS_ENDPOINT_URL
  • BEDROCK_EMBEDDING_MODEL
  • BEDROCK_MODEL_ID
  • MCP_TRANSPORT
  • MCP_HTTP_HOST
  • MCP_HTTP_PORT
  • MCP_SERVER_URL
  • VECTORSTORE_TABLE_NAME
  • VECTORSTORE_EMBEDDING_DIMENSION
  • HYBRID_SEMANTIC_WEIGHT
  • HYBRID_KEYWORD_WEIGHT
  • HYBRID_RRF_K
  • QUERY_CACHE_SIZE
  • HEALTH_CHECK_PORT
  • HEALTH_CHECK_HOST
  • LOG_LEVEL
  • LOG_FORMAT

Security Notes

The Redshift metadata tools (`list_tables`, `describe_table`) in `src/combined_mcp_server/redshift/tools.py` construct SQL queries using f-strings with schema/table names. If these inputs are derived from untrusted user input without proper validation or sanitization, this could lead to SQL injection vulnerabilities. The `run_query` tool itself uses parameterized queries for the main SQL input, which is safer.

Similar Servers

Stats

Interest Score0
Security Score6
Cost ClassMedium
Avg Tokens2000
Stars0
Forks0
Last Update2026-01-17

Tags

MCPRedshiftKnowledgebaseAWSVector StorePostgreSQLAI AgentLlamaIndex