fix: update reasoningEffort logic for gpt-5.1 models in SessionPrompt-ensureTitle

[WAY29]

No description provided.

[WAY29]

Ref: #4451

[rekram1-node]

@WAY29 do we need that fallback to none? I think just the gpt-5.1 case should do it right

[rekram1-node]

Im just gonna use the immediate fix im not sure that default would work across the board? So ill just do this for now

[rekram1-node]

/oc read the changes made in this pr, make this override existing change:

 if (small.providerID === "openai" || small.modelID.includes("gpt-5")) {
      if (small.modelID.includes("5.1")) {
        options["reasoningEffort"] = "low"
      } else {
        options["reasoningEffort"] = "minimal"
      }
    }

[opencode-agent]

The override has been implemented. The new logic now:

1. Checks if the provider is OpenAI or the model includes "gpt-5"
2. If the model includes "5.1", sets reasoningEffort to "low"
3. Otherwise, sets reasoningEffort to "minimal"

This simplifies the logic by removing the fallback to "none" and the specific "-5-" pattern matching, as requested.

opencode session  |  github run

[WAY29]

Im just gonna use the immediate fix im not sure that default would work across the board? So ill just do this for now

OK, fallback to None was a design choice I made with compatibility in mind for the future gpt model; this solution is also applicable now.