feat(token-viz): Phase 1 — token budget panel with color bar and breakdown table

ai_client.py

@@ -1715,6 +1715,19 @@ def send(
   else:
    raise ValueError(f"Unknown provider: {_provider}")
 
+def _add_bleed_derived(d: dict[str, Any], sys_tok: int = 0, tool_tok: int = 0) -> dict[str, Any]:
+ cur = d.get("current", 0)
+ lim = d.get("limit", 0)
+ d["estimated_prompt_tokens"] = cur
+ d["max_prompt_tokens"] = lim
+ d["utilization_pct"] = d.get("percentage", 0.0)
+ d["headroom_tokens"] = max(0, lim - cur)
+ d["would_trim"] = (lim - cur) < 20000
+ d["system_tokens"] = sys_tok
+ d["tools_tokens"] = tool_tok
+ d["history_tokens"] = max(0, cur - sys_tok - tool_tok)
+ return d
+
 def get_history_bleed_stats(md_content: str | None = None) -> dict[str, Any]:
  """
     Calculates how close the current conversation history is to the token limit.
@@ -1724,17 +1737,19 @@ def get_history_bleed_stats(md_content: str | None = None) -> dict[str, Any]:
  # For Anthropic, we have a robust estimator
   with _anthropic_history_lock:
    history_snapshot = list(_anthropic_history)
+  hist_only = _estimate_prompt_tokens([], history_snapshot) - 2500  # subtract fixed tools
+  sys_tok = max(1, int(len(md_content) / _CHARS_PER_TOKEN)) if md_content else 0
   current_tokens = _estimate_prompt_tokens([], history_snapshot)
   if md_content:
    current_tokens += max(1, int(len(md_content) / _CHARS_PER_TOKEN))
   limit_tokens = _ANTHROPIC_MAX_PROMPT_TOKENS
   percentage = (current_tokens / limit_tokens) * 100 if limit_tokens > 0 else 0
-  return {
+  return _add_bleed_derived({
    "provider": "anthropic",
    "limit": limit_tokens,
    "current": current_tokens,
    "percentage": percentage,
-  }
+  }, sys_tok=sys_tok, tool_tok=2500)
  elif _provider == "gemini":
   effective_limit = _history_trunc_limit if _history_trunc_limit > 0 else _GEMINI_MAX_INPUT_TOKENS
   if _gemini_chat:
@@ -1751,24 +1766,24 @@ def get_history_bleed_stats(md_content: str | None = None) -> dict[str, Any]:
     # Prepend context as a user part for counting
      history.insert(0, types.Content(role="user", parts=[types.Part.from_text(text=md_content)]))
     if not history:
-     return {
+     return _add_bleed_derived({
       "provider": "gemini",
       "limit": effective_limit,
       "current": 0,
       "percentage": 0,
-     }
+     })
     resp = _gemini_client.models.count_tokens(
      model=_model,
      contents=history
     )
     current_tokens = resp.total_tokens
     percentage = (current_tokens / effective_limit) * 100 if effective_limit > 0 else 0
-    return {
+    return _add_bleed_derived({
      "provider": "gemini",
      "limit": effective_limit,
      "current": current_tokens,
      "percentage": percentage,
-    }
+    }, sys_tok=0, tool_tok=0)
    except Exception as e:
     pass
   elif md_content:
@@ -1780,20 +1795,20 @@ def get_history_bleed_stats(md_content: str | None = None) -> dict[str, Any]:
     )
     current_tokens = resp.total_tokens
     percentage = (current_tokens / effective_limit) * 100 if effective_limit > 0 else 0
-    return {
+    return _add_bleed_derived({
      "provider": "gemini",
      "limit": effective_limit,
      "current": current_tokens,
      "percentage": percentage,
-    }
+    })
    except Exception as e:
     pass
-  return {
+  return _add_bleed_derived({
    "provider": "gemini",
    "limit": effective_limit,
    "current": 0,
    "percentage": 0,
-  }
+  })
  elif _provider == "gemini_cli":
   effective_limit = _history_trunc_limit if _history_trunc_limit > 0 else _GEMINI_MAX_INPUT_TOKENS
   limit_tokens = effective_limit
@@ -1802,12 +1817,12 @@ def get_history_bleed_stats(md_content: str | None = None) -> dict[str, Any]:
    u = _gemini_cli_adapter.last_usage
    current_tokens = u.get("input_tokens") or u.get("input", 0)
   percentage = (current_tokens / limit_tokens) * 100 if limit_tokens > 0 else 0
-  return {
+  return _add_bleed_derived({
    "provider": "gemini_cli",
    "limit": limit_tokens,
    "current": current_tokens,
    "percentage": percentage,
-  }
+  })
  elif _provider == "deepseek":
   limit_tokens = 64000
   current_tokens = 0
@@ -1829,15 +1844,15 @@ def get_history_bleed_stats(md_content: str | None = None) -> dict[str, Any]:
   if md_content: current_tokens += len(md_content)
   current_tokens = max(1, int(current_tokens / _CHARS_PER_TOKEN))
   percentage = (current_tokens / limit_tokens) * 100 if limit_tokens > 0 else 0
-  return {
+  return _add_bleed_derived({
    "provider": "deepseek",
    "limit": limit_tokens,
    "current": current_tokens,
    "percentage": percentage,
-  }
- return {
+  })
+ return _add_bleed_derived({
   "provider": _provider,
   "limit": 0,
   "current": 0,
   "percentage": 0,
- }
+ })