refactor: streamline token counting and message shaping logic

crates/zesdex-backend/src/app/runtime/actions/turn.rs

@@ -13,6 +13,7 @@ use sha2::Digest;
 use zesdex_cms::domain::repository::EditLogRepository;
 
 use crate::app::guard::Verdict;
+use crate::app::runtime::context::tokens::count_tokens;
 use crate::app::state::runtime::TurnEvent;
 use zesdex_cms::domain::repository::MemoryRepository;
 use crate::dto::chat::message::ChatMessage;
@@ -326,12 +327,11 @@ pub(super) fn run_agent_turn(
     let mut todo_retry_count = 0usize;
 
     loop {
-        let total_chars: usize = msgs
+        let token_estimate: usize = msgs
             .iter()
             .filter_map(|m| m.content.as_deref())
-            .map(str::len)
+            .map(count_tokens)
             .sum();
-        let token_estimate = total_chars / 4;
         let max_wire_tokens = tc.context_window;
 
         // Skip message compaction if abort was requested — the non-streaming
@@ -352,6 +352,7 @@ pub(super) fn run_agent_turn(
                     max_wire_tokens,
                     false,
                     Some(&tc.client),
+                    Some(&tc.abort_flag),
                 );
 
             // Dispatch the compacted messages to the main thread so the local session history
@@ -488,12 +489,12 @@ pub(super) fn run_agent_turn(
 
         let (mut tok_in, mut tok_out) = final_usage.unwrap_or((0, 0));
         if tok_in == 0 {
-            let total_chars: usize = wire_msgs
+            let total_tokens: usize = wire_msgs
                 .iter()
                 .filter_map(|m| m.content.as_deref())
-                .map(str::len)
+                .map(count_tokens)
                 .sum();
-            tok_in = (total_chars / 4).max(1) as u64;
+            tok_in = total_tokens.max(1) as u64;
         }
         if tok_out == 0 {
             let response_chars = response.content.as_deref().map_or(0, str::len);