Parallelize fragment connections inside list_mcp_tools so all fragments are handled concurrently

[HoaiGit]

Describe the Problem

Saw some changes being made for list_mcp_tools:

• https://github.com/SAP/cloud-sdk-python/releases/tag/v0.41.0
• Support MCP tools caching #178

From my understanding list_mcp_tools() fetches the registered MCP fragments sequentially.
That means fetch time scales with the number of such fragments.

N fragments x ~3s ( equals ~44 seconds for N=14 for example)

Propose a Solution

Parallelizing these fragment fetches could further improve performance, especially when there is no cache yet.
Fetch time would not scale with the number of MCP fragments.

Describe Alternatives

No response

Affected Development Phase

Getting Started

Impact

No Impact

Timeline

No response