bitDive/mcp-server

JDK 17+ and Maven required
./mvnw clean package && ./mvnw spring-boot:run

Usage

• •Connect an MCP client to the SSE endpoint http://localhost:8089/sse and message endpoint http://localhost:8089/mcp/message
• •For BitDive Cloud add {"mcpServers":{"bitdive":{"url":"https://cloud.bitdive.io/mcp/sse","name":"BitDive MCP Server"}}}
• •Obtain an API token from the BitDive dashboard MCP Integration section and add a user rule passing it to the AI

Key features

• ✓Trace tools: findTraceAll, findTraceForMethod, findTraceForMethodBetweenTime
• ✓Monitoring tools: getCurrentHeapMapAllSystem, per-module/service/class heap maps, getLastCallService
• ✓Spring Boot 3.2.0 with Spring AI 1.0.0 MCP Server WebFlux
• ✓PostgreSQL storage with HashiCorp Vault and Bouncy Castle security

Best for

AI assistants querying BitDive APM traces and performance heat maps.

Caveats

• ⚠Current authentication requires passing the API key via prompts/rules; one-click integration is not yet available
• ⚠Requires a running BitDive platform instance (self-hosted infrastructure or cloud)
• ⚠Self-hosting needs configuration of POSTGRES_URL, POSTGRES_USER, POSTGRES_PASS, VAULT_URL and TOKEN_SECRET