Add AutoTune config to prefix scorer to make it explicit when auto vs. manual config is used

[liu-cong]

Previously the prefix cache scorer implicitly overrides the BlockSize and LRUCapacityPerServer via model server metrics. This is the default behavior we want. However it can be confusing if users manually configure these parameters.

This PR adds an autoTune flag to the prefix cache scorer config to allow the user to explicitly set whether they want autoTune or not, and is defaulted to true.

This also allows use to manually set the CPU cache capacity because the corresponding metric isn't ready in vllm yet.

Test

Tried the following config:

 pluginsConfigFile: "default-plugins.yaml"
  # This is the plugins configuration file.
  pluginsCustomConfig:
    default-plugins.yaml: |
      apiVersion: inference.networking.x-k8s.io/v1alpha1
      kind: EndpointPickerConfig
      plugins:
      - type: queue-scorer
      - type: kv-cache-utilization-scorer
      - type: prefix-cache-scorer
        name: gpu-prefix-cache-scorer
        parameters:
          autoTune: true
      - type: prefix-cache-scorer
        name: cpu-prefix-cache-scorer
        parameters:
          autoTune: false
          lruCapacityPerServer: 41000
      schedulingProfiles:
      - name: default
        plugins:
        - pluginRef: queue-scorer
          weight: 2
        - pluginRef: kv-cache-utilization-scorer
          weight: 2
        - pluginRef: gpu-prefix-cache-scorer
          weight: 2
        - pluginRef: cpu-prefix-cache-scorer
          weight: 1

And got the following log:

{"level":"Level(-2)","ts":"2025-11-20T22:58:56Z","caller":"prefix/plugin.go:190","msg":"PrefixCachePlugin initialized","config":{"autoTune":true,"blockSize":64,"maxPrefixBlocksToMatch":256,"lruCapacityPerServer":31250}}                       │
│ {"level":"Level(-2)","ts":"2025-11-20T22:58:56Z","caller":"prefix/plugin.go:190","msg":"PrefixCachePlugin initialized","config":{"autoTune":false,"blockSize":64,"maxPrefixBlocksToMatch":256,"lruCapacityPerServer":41000}}

[k8s-ci-robot]

Skipping CI for Draft Pull Request.
If you want CI signal for your change, please convert it to an actual PR.
You can still manually trigger a test run with /test all

[netlify]

✅ Deploy Preview for gateway-api-inference-extension ready!

Name	Link
🔨 Latest commit	23d4d96
🔍 Latest deploy log	https://app.netlify.com/projects/gateway-api-inference-extension/deploys/691f99ad09f8320008d73c90
😎 Deploy Preview	https://deploy-preview-1888--gateway-api-inference-extension.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify project configuration.

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: liu-cong

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• pkg/epp/scheduling/framework/plugins/multi/prefix/OWNERS [liu-cong]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[zetxqx]

/lgtm

[liu-cong]

/hold

Running some test

[zetxqx]

/lgtm

[liu-cong]

/hold cancel