feat(config): add per-request token budget

docs/advanced/configuration.mdx

@@ -31,6 +31,11 @@ Configure Strix using environment variables or a config file.
   Request timeout in seconds for LLM calls.
 </ParamField>
 
+<ParamField path="STRIX_LLM_MAX_TOKENS" type="integer">
+  Optional maximum output tokens for each agent LLM request. Leave unset to use
+  the provider/model default.
+</ParamField>
+
 <ParamField path="STRIX_LLM_MAX_RETRIES" default="5" type="integer">
   Maximum number of retries for LLM API calls on transient failures.
 </ParamField>

strix/config/settings.py

@@ -52,6 +52,7 @@ class LlmSettings(BaseSettings):
         default=False,
         alias="LLM_DISABLE_STREAMING",
     )
+    max_tokens: int | None = Field(default=None, gt=0, alias="STRIX_LLM_MAX_TOKENS")
     timeout: int = Field(default=300, alias="LLM_TIMEOUT")
 
 

strix/core/inputs.py

@@ -133,11 +133,13 @@ def make_model_settings(
     request_timeout: float | None = None,
     prompt_cache: bool = True,
     extra_headers: dict[str, str] | None = None,
+    max_tokens: int | None = None,
 ) -> ModelSettings:
     model_settings = ModelSettings(
         parallel_tool_calls=False,
         retry=DEFAULT_MODEL_RETRY,
         include_usage=True,
+        max_tokens=max_tokens,
         extra_args=request_timeout_extra_args(request_timeout),
         extra_headers=dict(extra_headers) if extra_headers else None,
     )

strix/core/runner.py

@@ -251,6 +251,7 @@ async def _spill_to_workspace(output_id: str, text: str) -> str | None:
             request_timeout=settings.llm.timeout,
             prompt_cache=settings.llm.prompt_cache,
             extra_headers=settings.llm.extra_headers,
+            max_tokens=settings.llm.max_tokens,
         )
         run_config = RunConfig(
             model=resolved_model,

tests/test_config_loader.py

@@ -10,7 +10,7 @@
 from pydantic.fields import FieldInfo
 
 from strix.config import loader
-from strix.config.settings import ContextSettings
+from strix.config.settings import ContextSettings, LlmSettings
 
 
 if TYPE_CHECKING:
@@ -28,6 +28,7 @@
     "OLLAMA_API_BASE",
     "STRIX_REASONING_EFFORT",
     "STRIX_FORCE_REQUIRED_TOOL_CHOICE",
+    "STRIX_LLM_MAX_TOKENS",
     "LLM_TIMEOUT",
     "PERPLEXITY_API_KEY",
     # RuntimeSettings
@@ -130,6 +131,10 @@ def test_tool_output_max_bytes_accepts_floor() -> None:
     assert ContextSettings(STRIX_TOOL_OUTPUT_MAX_BYTES=1024).tool_output_max_bytes == 1024
 
 
+def test_llm_max_tokens_env_alias() -> None:
+    assert LlmSettings(STRIX_LLM_MAX_TOKENS=12_000).max_tokens == 12_000
+
+
 # --------------------------------------------------------------------------- #
 # _aliases_for
 # --------------------------------------------------------------------------- #

tests/test_inputs.py

@@ -266,6 +266,16 @@ def test_make_model_settings_sets_request_timeout() -> None:
     assert settings.extra_args["timeout"] == 300.0
 
 
+def test_make_model_settings_sets_configured_token_budget() -> None:
+    settings = make_model_settings(
+        "none",
+        model_name="gpt-4o",
+        max_tokens=12_000,
+    )
+
+    assert settings.max_tokens == 12_000
+
+
 def test_make_model_settings_omits_timeout_when_unset() -> None:
     settings = make_model_settings("none", model_name="gpt-4o")
 

tests/test_runner_rate_limit.py

@@ -41,6 +41,7 @@ async def test_persistent_rate_limit_stops_gracefully(
             timeout=300,
             prompt_cache=True,
             extra_headers=None,
+            max_tokens=None,
         ),
         runtime=types.SimpleNamespace(max_context_images=3),
     )

tests/test_runner_root_prompt.py

@@ -49,6 +49,7 @@ def _patch_engine_scaffold(
             timeout=300,
             prompt_cache=True,
             extra_headers=None,
+            max_tokens=12_000,
         ),
         runtime=types.SimpleNamespace(max_context_images=3),
     )
@@ -74,10 +75,15 @@ async def _cleanup(*_args: Any, **_kwargs: Any) -> None:
 
     monkeypatch.setattr(runner, "build_root_task", lambda _scan_config: "task")
     monkeypatch.setattr(runner, "build_scope_context", lambda _scan_config: scope_context)
-    monkeypatch.setattr(runner, "make_model_settings", lambda *_args, **_kwargs: object())
 
     captured: dict[str, Any] = {}
 
+    def _make_model_settings(*_args: Any, **kwargs: Any) -> object:
+        captured["model_settings_kwargs"] = kwargs
+        return object()
+
+    monkeypatch.setattr(runner, "make_model_settings", _make_model_settings)
+
     def _build_strix_agent(**kwargs: Any) -> object:
         if kwargs.get("is_root") and "kwargs" not in captured:
             captured["kwargs"] = kwargs
@@ -176,3 +182,20 @@ async def test_root_prompt_options_default_to_none(
     kwargs = captured["kwargs"]
     assert kwargs["instructions_override"] is None
     assert kwargs["system_prompt_context"] == {"scope": "built-in"}
+
+
+@pytest.mark.asyncio
+async def test_llm_max_tokens_flows_into_model_settings(
+    monkeypatch: pytest.MonkeyPatch,
+    tmp_path: Any,
+) -> None:
+    captured = _patch_engine_scaffold(monkeypatch, tmp_path, {"scope": "built-in"})
+
+    await runner.run_strix_scan(
+        scan_config={"targets": [], "scan_mode": "deep"},
+        scan_id="scan-token-budget",
+        image="img",
+        coordinator=AgentCoordinator(),
+    )
+
+    assert captured["model_settings_kwargs"]["max_tokens"] == 12_000