### Check for existing issues

- [x] I have searched the existing issues and checked that my issue is not a duplicate.

### What happened?

## What Happened

After upgrading LiteLLM Proxy from a previous version to **1.88.0**, the Prometheus `/metrics` endpoint returns empty data when scraped by Prometheus or Grafana Alloy.

The root cause is that `/metrics` (without trailing slash) now responds with a **307 Temporary Redirect** to `/metrics/` (with trailing slash). Prometheus-compatible scrapers do not follow redirects by design, so they receive an empty body and silently record no metrics — while the `up` metric still shows `1`, making the issue hard to detect.

Accessing `/metrics/` (with trailing slash) directly returns the correct Prometheus metrics as expected.

This appears to be an unintentional breaking change introduced in 1.88.0, where the route registration changed from `/metrics` to `/metrics/`.

### Steps to Reproduce

## Steps to Reproduce

1. Deploy LiteLLM Proxy 1.88.0 with the following `config.yaml`: ``` litellm_settings: callbacks: ["prometheus"] ``` 2. Configure Grafana Alloy (or any Prometheus scraper) with: ``` prometheus.scrape "litellm" { targets …