docs: add governed endpoint benchmark example

[cerebrixos]

This adds a docs example for benchmarking a governed OpenAI-compatible endpoint with GuideLLM.

Why this felt useful here:

• GuideLLM already targets OpenAI-compatible endpoints through openai_http
• many teams benchmark a raw model server, but deploy a more complex production path
• this example shows how to measure the user-visible latency and throughput of the full governed path without changing GuideLLMs normal backend model

The concrete example uses Tuning Engines, but the workflow applies to any OpenAI-compatible control plane or gateway.

---

git log

commit c97fdd7
Author: cerebrixos 248534569+cerebrixos@users.noreply.github.com
Date: Sat Jun 27 10:49:19 2026 -0400

docs: add governed endpoint benchmark example

[mergify]

Hi @cerebrixos, the DCO check has failed. Please click on DCO in the Checks section for instructions on how to resolve this.

[cerebrixos]

I opened #869 as a signed-off replacement for this PR so the DCO check can pass cleanly. Closing this one to keep review on the compliant branch.