"DHL" matching MCP servers:

Enables context-aware memory and multi-model orchestration through tools for storing/retrieving dialog turns, searching memory, and routing to AI providers based on task types.

MCP server for Threadline — persistent memory and context layer for AI agents. inject() before your LLM call, update() after. Relevance-scored injection, grant-based access, user-owned context.

Bridges AI agents with watsonx.data document libraries, enabling natural language queries and intelligent retrieval from document libraries.

A self-hosted MCP server that enables a voice assistant to answer a small business owner's weekly cash questions — bank balances, runway and 13-week forecasts, duplicate or unusual charges, affordability of new hires, spending changes, and importable books — using ten tools backed by deterministic calculations and SQLite memory that persists alerts, reviewed charges, and prior briefs across sessions.