refactor(token): simplify token tracking and status display

- Remove redundant token breakdown in session summary output
- Streamline status command to show only essential token metrics
- Simplify subagent token tracking logic and logging
- Add error handling for token tracker status checks

Author: cellwebb

src/clippy/agent/token_tracker.py

@@ -29,7 +29,8 @@ def from_api_response(
         """Create TokenUsage from API response usage data.
 
         Args:
-            usage_data: Usage dict from API response with prompt_tokens, completion_tokens, total_tokens
+            usage_data: Usage dict from API response with prompt_tokens, completion_tokens,
+                      total_tokens
             operation: Type of operation (main_agent, subagent, etc.)
             operation_id: Identifier for the specific operation
             model: Model name used
@@ -155,18 +156,6 @@ def get_summary(self) -> dict[str, Any]:
                 "completion_tokens": grand_total.completion_tokens,
                 "total_tokens": grand_total.total_tokens,
             },
-            "main_agent": {
-                "prompt_tokens": self.main_agent.prompt_tokens,
-                "completion_tokens": self.main_agent.completion_tokens,
-                "total_tokens": self.main_agent.total_tokens,
-            },
-            "subagents": {
-                "count": len(self.subagents),
-                "total_tokens": subagent_total.total_tokens,
-                "prompt_tokens": subagent_total.prompt_tokens,
-                "completion_tokens": subagent_total.completion_tokens,
-                "details": subagent_details,
-            },
         }
 
     def is_enabled(self) -> bool:

src/clippy/cli/commands/system.py

@@ -50,7 +50,7 @@ def handle_help_command(console: Console) -> CommandResult:
         "    Examples: /model add (wizard) or /model add claude-code claude-4-5 --name sonnet\n"
         "  /model remove <name> - Remove a saved model\n"
         "  /model set-default <name> - Set model as default (permanent)\n"
-        "  /model threshold <name> <tokens> - Set compaction threshold\n"
+        "  /model threshold <name> <tokens> - Set compaction threshold\n\n"
         "[bold]Subagent Configuration:[/bold]\n"
         "  /subagent list - Show subagent type configurations\n"
         "  /subagent set <type> <model> - Set model for a subagent type\n"
@@ -174,19 +174,9 @@ def handle_status_command(agent: ClippyAgent, console: Console) -> CommandResult
             note = "[dim]Note: Usage % is estimated for ~128k context window[/dim]"
 
         # Build actual usage summary (from API)
+        actual_prompt = session_summary["total"]["prompt_tokens"]
+        actual_completion = session_summary["total"]["completion_tokens"]
         actual_tokens = session_summary["total"]["total_tokens"]
-        actual_main = session_summary["main_agent"]["total_tokens"]
-        actual_subagents = session_summary["subagents"]["total_tokens"]
-        actual_subagent_count = session_summary["subagents"]["count"]
-
-        # Calculate estimated cost (using approximate GPT-4 rates)
-        prompt_cost = (
-            session_summary["total"]["prompt_tokens"] * 0.00003
-        )  # $0.03 per 1K prompt tokens
-        completion_cost = (
-            session_summary["total"]["completion_tokens"] * 0.00006
-        )  # $0.06 per 1K completion tokens
-        estimated_cost = prompt_cost + completion_cost
 
         # Build status content
         status_content = (
@@ -203,22 +193,19 @@ def handle_status_command(agent: ClippyAgent, console: Console) -> CommandResult
         if actual_tokens > 0:
             status_content += (
                 f"[bold]Actual API Usage:[/bold]\n"
-                f"  Total: [cyan]{actual_tokens:,}[/cyan] tokens\n"
-                f"  Main Agent: [blue]{actual_main:,}[/blue] tokens\n"
-                f"  Subagents: [green]{actual_subagents:,}[/green] tokens "
-                f"([dim]{actual_subagent_count} subagents[/dim])\n"
+                f"  Input:  [cyan]{actual_prompt:,}[/cyan] tokens\n"
+                f"  Output: [cyan]{actual_completion:,}[/cyan] tokens\n"
+                f"  Total:  [cyan]{actual_tokens:,}[/cyan] tokens\n\n"
             )
 
-            if estimated_cost > 0:
-                status_content += f"  💰 Est. Cost: [yellow]${estimated_cost:.4f}[/yellow]\n"
-
-            status_content += "\n"
-
         status_content += f"[bold]Message Breakdown:[/bold]\n    {message_breakdown}\n\n{note}"
 
         # Show token tracking status if disabled
-        if not tracker.is_enabled():
-            status_content += "\n\n[dim]⚠️ Token tracking is disabled[/dim]"
+        try:
+            if not tracker.is_enabled():
+                status_content += "\n\n[dim]⚠️ Token tracking is disabled[/dim]"
+        except Exception:
+            pass
 
         console.print(
             Panel.fit(

src/clippy/tools/delegate_to_subagent.py

@@ -185,7 +185,8 @@ def create_subagent_and_execute(
                 result.actual_token_usage, name, result.metadata.get("model", "")
             )
             logger.info(
-                f"Tracked {result.actual_token_usage.get('total_tokens', 0)} tokens for subagent '{name}'"
+                f"Tracked {result.actual_token_usage.get('total_tokens', 0)} tokens "
+                f"for subagent '{name}'"
             )
 
         if result.success: