Enhance LLM configuration with thinking budget support

- Added optional `thinking_budget` parameter to `config.yaml` for Gemini model, allowing for reserved thinking tokens.
- Updated `LlmSettings` class in `config.py` to include `thinking_budget` and adjusted parsing logic accordingly.
- Modified LLM call logic in `llm.py` to utilize `thinking_budget` for calculating `max_output_tokens`, improving output management.

config.yaml

@@ -2,6 +2,9 @@
 # Two independent profiles, one per pipeline stage. Each accepts the same keys:
 #   provider: openai | anthropic | gemini
 #   openai_model / anthropic_model / gemini_model / max_tokens / temperature
+#   thinking_budget (optional, Gemini only): reserved thinking tokens. When set,
+#   Gemini's max_output_tokens is max_tokens + thinking_budget, so max_tokens
+#   is the visible-output budget. Omit it to keep max_tokens as the shared cap.
 llm:
   # Part_One: vector-search query synthesis (small structured JSON in,
   # 1-2 short queries out). A lighter/cheaper model is sufficient here.
@@ -22,7 +25,8 @@ llm:
     openai_model: gpt-4o
     anthropic_model: claude-opus-4-5
     gemini_model: gemini-2.5-pro
-    max_tokens: 16384
+    max_tokens: 4096
+    thinking_budget: 9216
     temperature: 0.0
 
 # ── Crop-first worklist ────────────────────────────────────────────────────────

pipeline/config.py

@@ -20,6 +20,7 @@ class LlmSettings:
     gemini_model: str
     max_tokens: int
     temperature: float
+    thinking_budget: int | None
 
 
 @dataclass(frozen=True)
@@ -119,6 +120,7 @@ class Settings:
 
 def _parse_llm_settings(raw: dict, defaults: dict) -> LlmSettings:
     """Build an LlmSettings profile from a config sub-block, falling back to defaults."""
+    thinking_raw = raw["thinking_budget"] if "thinking_budget" in raw else defaults.get("thinking_budget")
     return LlmSettings(
         provider=str(raw.get("provider", defaults["provider"])).strip().lower(),
         openai_model=str(raw.get("openai_model", defaults["openai_model"])),
@@ -126,6 +128,7 @@ def _parse_llm_settings(raw: dict, defaults: dict) -> LlmSettings:
         gemini_model=str(raw.get("gemini_model", defaults["gemini_model"])),
         max_tokens=int(raw.get("max_tokens", defaults["max_tokens"])),
         temperature=float(raw.get("temperature", defaults["temperature"])),
+        thinking_budget=None if thinking_raw is None else int(thinking_raw),
     )
 
 
@@ -235,14 +238,16 @@ def load_settings(
         "gemini_model": "gemini-2.5-flash",
         "max_tokens": 8192,
         "temperature": 0.0,
+        "thinking_budget": None,
     }
     advice_generation_defaults = {
         "provider": "gemini",
         "openai_model": "gpt-4o",
         "anthropic_model": "claude-opus-4-5",
         "gemini_model": "gemini-2.5-pro",
-        "max_tokens": 65536,
+        "max_tokens": 8192,
         "temperature": 0.0,
+        "thinking_budget": 8192,
     }
 
     return Settings(

pipeline/stages/llm.py

@@ -101,7 +101,12 @@ def call_llm(
     """Dispatch a system/user prompt pair to the provider configured in llm_cfg."""
     provider = llm_cfg.provider
 
-    logger.info("Calling LLM provider=%s model=%s", provider, _model_name(llm_cfg))
+    logger.info(
+        "Calling LLM provider=%s model=%s thinking_budget=%s",
+        provider,
+        _model_name(llm_cfg),
+        llm_cfg.thinking_budget,
+    )
     logger.debug(
         "LLM REQUEST (system_prompt + JSON user content):\n--- SYSTEM PROMPT ---\n%s\n--- USER JSON ---\n%s",
         system_prompt,
@@ -140,6 +145,12 @@ def _call_gemini(
     from google import genai
     from google.genai import types
 
+    max_output_tokens = llm_cfg.max_tokens
+    thinking_config = None
+    if llm_cfg.thinking_budget is not None:
+        max_output_tokens = llm_cfg.max_tokens + llm_cfg.thinking_budget
+        thinking_config = types.ThinkingConfig(thinking_budget=llm_cfg.thinking_budget)
+
     client = genai.Client(api_key=settings.gemini_api_key)
     response = client.models.generate_content(
         model=llm_cfg.gemini_model,
@@ -147,7 +158,8 @@ def _call_gemini(
         config=types.GenerateContentConfig(
             system_instruction=system_prompt,
             temperature=llm_cfg.temperature,
-            max_output_tokens=llm_cfg.max_tokens,
+            max_output_tokens=max_output_tokens,
+            thinking_config=thinking_config,
             response_mime_type="application/json" if json_output else None,
         ),
     )