Responses file_search wire up additional params

This adds passing of max_num_results from the file_search tool call
into the knowledge_search tool, as well as logs warnings if the
filters or ranking_options params are used since those are not wired
up yet. And, it adds the API surface for filters and ranking options
so we don't have to generate clients again as we add that.

Signed-off-by: Ben Browning <bbrownin@redhat.com>

docs/_static/llama-stack-spec.html

@@ -7255,6 +7255,35 @@
                             "type": "string"
                         }
                     },
+                    "filters": {
+                        "type": "object",
+                        "additionalProperties": {
+                            "oneOf": [
+                                {
+                                    "type": "null"
+                                },
+                                {
+                                    "type": "boolean"
+                                },
+                                {
+                                    "type": "number"
+                                },
+                                {
+                                    "type": "string"
+                                },
+                                {
+                                    "type": "array"
+                                },
+                                {
+                                    "type": "object"
+                                }
+                            ]
+                        }
+                    },
+                    "max_num_results": {
+                        "type": "integer",
+                        "default": 10
+                    },
                     "ranking_options": {
                         "type": "object",
                         "properties": {

docs/_static/llama-stack-spec.yaml

@@ -5157,6 +5157,19 @@ components:
           type: array
           items:
             type: string
+        filters:
+          type: object
+          additionalProperties:
+            oneOf:
+              - type: 'null'
+              - type: boolean
+              - type: number
+              - type: string
+              - type: array
+              - type: object
+        max_num_results:
+          type: integer
+          default: 10
         ranking_options:
           type: object
           properties:

llama_stack/apis/agents/openai_responses.py

@@ -409,8 +409,9 @@ class FileSearchRankingOptions(BaseModel):
 class OpenAIResponseInputToolFileSearch(BaseModel):
     type: Literal["file_search"] = "file_search"
     vector_store_ids: list[str]
+    filters: dict[str, Any] | None = None
+    max_num_results: int | None = Field(default=10, ge=1, le=50)
     ranking_options: FileSearchRankingOptions | None = None
-    # TODO: add filters, max_num_results
 
 
 class ApprovalFilter(BaseModel):

llama_stack/providers/inline/agents/meta_reference/openai_responses.py

@@ -64,7 +64,7 @@ from llama_stack.apis.inference.inference import (
     OpenAIToolMessageParam,
     OpenAIUserMessageParam,
 )
-from llama_stack.apis.tools.tools import ToolGroups, ToolRuntime
+from llama_stack.apis.tools import RAGQueryConfig, ToolGroups, ToolRuntime
 from llama_stack.log import get_logger
 from llama_stack.models.llama.datatypes import ToolDefinition, ToolParamDefinition
 from llama_stack.providers.utils.inference.openai_compat import convert_tooldef_to_openai_tool
@@ -699,7 +699,15 @@ class OpenAIResponsesImpl:
                         t for t in ctx.response_tools if isinstance(t, OpenAIResponseInputToolFileSearch)
                     )
                     if response_file_search_tool:
+                        if response_file_search_tool.filters:
+                            logger.warning("Filters are not yet supported for file_search tool")
+                        if response_file_search_tool.ranking_options:
+                            logger.warning("Ranking options are not yet supported for file_search tool")
                         tool_kwargs["vector_db_ids"] = response_file_search_tool.vector_store_ids
+                        tool_kwargs["query_config"] = RAGQueryConfig(
+                            mode="vector",
+                            max_chunks=response_file_search_tool.max_num_results,
+                        )
                 result = await self.tool_runtime_api.invoke_tool(
                     tool_name=function.name,
                     kwargs=tool_kwargs,