Stop sending prompt cache keys on Claude chat conversions (#2003)

Responses conversions still use promptCacheKey, but chat completions now stay a pure shape transform. This keeps Claude -> chat requests aligned with providers that do not understand the field and keeps stream checks consistent with production behavior.

Constraint: Issue #1919 requires removing prompt_cache_key from Claude -> OpenAI Chat requests
Rejected: Add a runtime toggle for chat injection | requested behavior is unconditional removal
Confidence: high
Scope-risk: narrow
Reversibility: clean
Directive: Keep promptCacheKey limited to Claude -> Responses conversions unless a provider-specific contract is proven
Tested: cargo test anthropic_to_openai
Tested: cargo test anthropic_to_responses_with_cache_key
Tested: cargo test transform_claude_request_for_api_format_responses
Not-tested: Full src-tauri test suite
Related: #1919

src-tauri/src/provider.rs

@@ -282,9 +282,9 @@ pub struct ProviderMeta {
     /// 是否将 base_url 视为完整 API 端点（不拼接 endpoint 路径）
     #[serde(rename = "isFullUrl", skip_serializing_if = "Option::is_none")]
     pub is_full_url: Option<bool>,
-    /// Prompt cache key for OpenAI-compatible endpoints.
-    /// When set, injected into converted requests to improve cache hit rate.
-    /// If not set, provider ID is used automatically during format conversion.
+    /// Prompt cache key for OpenAI Responses-compatible endpoints.
+    /// When set, injected into converted Responses requests to improve cache hit rate.
+    /// If not set, provider ID is used automatically during Claude -> Responses conversion.
     #[serde(rename = "promptCacheKey", skip_serializing_if = "Option::is_none")]
     pub prompt_cache_key: Option<String>,
     /// 累加模式应用中，该 provider 是否已写入 live config。