fix: cascade correlation — fallback to _latest MITM usage

When the MITM can't extract a cascade ID from the intercepted request
(Content-Length: 0 / chunked encoding), usage is stored under '_latest'.
Now usage_from_poll and completions try the exact cascade_id first,
then fall back to '_latest' so MITM-captured tokens are actually used.

src/api/completions.rs

@@ -308,8 +308,12 @@ async fn chat_completions_sync(
 ) -> axum::response::Response {
     let result = poll_for_response(&state, &cascade_id, timeout).await;
 
-    // Check MITM store first for real intercepted usage
+    // Check MITM store first for real intercepted usage (fallback to _latest)
-    let (prompt_tokens, completion_tokens, cached_tokens) = if let Some(mitm_usage) = state.mitm_store.take_usage(&cascade_id).await {
+    let mitm = match state.mitm_store.take_usage(&cascade_id).await {
+        Some(u) => Some(u),
+        None => state.mitm_store.take_usage("_latest").await,
+    };
+    let (prompt_tokens, completion_tokens, cached_tokens) = if let Some(mitm_usage) = mitm {
         (mitm_usage.input_tokens, mitm_usage.output_tokens, mitm_usage.cache_read_input_tokens)
     } else if let Some(u) = &result.usage {
         (u.input_tokens, u.output_tokens, 0)

src/api/responses.rs

@@ -277,7 +277,12 @@ async fn usage_from_poll(
     output_text: &str,
 ) -> Usage {
     // Priority 1: MITM intercepted data (most accurate — includes cache tokens)
-    if let Some(mitm_usage) = mitm_store.take_usage(cascade_id).await {
+    // Try exact cascade_id match first, then fall back to "_latest" (unmatched)
+    let mitm_usage = match mitm_store.take_usage(cascade_id).await {
+        Some(u) => Some(u),
+        None => mitm_store.take_usage("_latest").await,
+    };
+    if let Some(mitm_usage) = mitm_usage {
         tracing::debug!(
             input = mitm_usage.input_tokens,
             output = mitm_usage.output_tokens,