lint fix and rename file search

Signed-off-by: Francisco Javier Arceo <farceo@redhat.com>

llama_stack/core/routers/tool_runtime.py

@@ -41,7 +41,7 @@ class ToolRuntimeRouter(ToolRuntime):
             query_config: FileSearchConfig | None = None,
         ) -> FileSearchResult:
             logger.debug(f"ToolRuntimeRouter.FileSearchToolImpl.query: {vector_store_ids}")
-            provider = await self.routing_table.get_provider_impl("knowledge_search")
+            provider = await self.routing_table.get_provider_impl("file_search")
             return await provider.query(content, vector_store_ids, query_config)
 
         async def insert(

llama_stack/core/routing_tables/toolgroups.py

@@ -18,7 +18,7 @@ logger = get_logger(name=__name__, category="core::routing_tables")
 
 
 def parse_toolgroup_from_toolgroup_name_pair(toolgroup_name_with_maybe_tool_name: str) -> str | None:
-    # handle the funny case like "builtin::file_search/knowledge_search"
+    # handle the funny case like "builtin::file_search/file_search"
     parts = toolgroup_name_with_maybe_tool_name.split("/")
     if len(parts) == 2:
         return parts[0]

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

@@ -84,7 +84,7 @@ from .persistence import AgentPersistence
 from .safety import SafetyException, ShieldRunnerMixin
 
 TOOLS_ATTACHMENT_KEY_REGEX = re.compile(r"__tools_attachment__=(\{.*?\})")
-MEMORY_QUERY_TOOL = "knowledge_search"
+MEMORY_QUERY_TOOL = "file_search"
 WEB_SEARCH_TOOL = "web_search"
 RAG_TOOL_GROUP = "builtin::file_search"
 
@@ -927,7 +927,7 @@ class ChatAgent(ShieldRunnerMixin):
         """Parse a toolgroup name into its components.
 
         Args:
-            toolgroup_name: The toolgroup name to parse (e.g. "builtin::file_search/knowledge_search")
+            toolgroup_name: The toolgroup name to parse (e.g. "builtin::file_search/file_search")
 
         Returns:
             A tuple of (tool_type, tool_group, tool_name)

llama_stack/providers/inline/agents/meta_reference/responses/streaming.py

@@ -675,7 +675,7 @@ class StreamingResponseOrchestrator:
                             # Emit output_item.added event for the new function call
                             self.sequence_number += 1
                             is_mcp_tool = tool_call.function.name and tool_call.function.name in self.mcp_tool_to_server
-                            if not is_mcp_tool and tool_call.function.name not in ["web_search", "knowledge_search"]:
+                            if not is_mcp_tool and tool_call.function.name not in ["web_search", "file_search"]:
                                 # for MCP tools (and even other non-function tools) we emit an output message item later
                                 function_call_item = OpenAIResponseOutputMessageFunctionToolCall(
                                     arguments="",  # Will be filled incrementally via delta events
@@ -900,7 +900,7 @@ class StreamingResponseOrchestrator:
                     id=matching_item_id,
                     status="in_progress",
                 )
-            elif tool_call.function.name == "knowledge_search":
+            elif tool_call.function.name == "file_search":
                 item = OpenAIResponseOutputMessageFileSearchToolCall(
                     id=matching_item_id,
                     status="in_progress",
@@ -1019,7 +1019,7 @@ class StreamingResponseOrchestrator:
                     raise ValueError(f"Tool {tool_name} not found")
                 self.ctx.chat_tools.append(make_openai_tool(tool_name, tool))
             elif input_tool.type == "file_search":
-                tool_name = "knowledge_search"
+                tool_name = "file_search"
                 tool = await self.tool_executor.tool_groups_api.get_tool(tool_name)
                 if not tool:
                     raise ValueError(f"Tool {tool_name} not found")

llama_stack/providers/inline/agents/meta_reference/responses/tool_executor.py

@@ -104,12 +104,12 @@ class ToolExecutor:
             citation_files=result.metadata.get("citation_files") if result and result.metadata else None,
         )
 
-    async def _execute_knowledge_search_via_vector_store(
+    async def _execute_file_search_via_vector_store(
         self,
         query: str,
         response_file_search_tool: OpenAIResponseInputToolFileSearch,
     ) -> ToolInvocationResult:
-        """Execute knowledge search using vector_stores.search API with filters support."""
+        """Execute file search using vector_stores.search API with filters support."""
         search_results = []
 
         # Create search tasks for all vector stores
@@ -139,7 +139,7 @@ class ToolExecutor:
         content_items = []
         content_items.append(
             TextContentItem(
-                text=f"knowledge_search tool found {len(search_results)} chunks:\nBEGIN of knowledge_search tool results.\n"
+                text=f"file_search tool found {len(search_results)} chunks:\nBEGIN of file_search tool results.\n"
             )
         )
 
@@ -158,7 +158,7 @@ class ToolExecutor:
             content_items.append(TextContentItem(text=text_content))
             unique_files.add(file_id)
 
-        content_items.append(TextContentItem(text="END of knowledge_search tool results.\n"))
+        content_items.append(TextContentItem(text="END of file_search tool results.\n"))
 
         citation_instruction = ""
         if unique_files:
@@ -224,7 +224,7 @@ class ToolExecutor:
                 output_index=output_index,
                 sequence_number=sequence_number,
             )
-        elif function_name == "knowledge_search":
+        elif function_name == "file_search":
             sequence_number += 1
             progress_event = OpenAIResponseObjectStreamResponseFileSearchCallInProgress(
                 item_id=item_id,
@@ -246,7 +246,7 @@ class ToolExecutor:
             yield ToolExecutionResult(stream_event=searching_event, sequence_number=sequence_number)
 
         # For file search, emit searching event
-        if function_name == "knowledge_search":
+        if function_name == "file_search":
             sequence_number += 1
             searching_event = OpenAIResponseObjectStreamResponseFileSearchCallSearching(
                 item_id=item_id,
@@ -283,17 +283,17 @@ class ToolExecutor:
                         tool_name=function_name,
                         kwargs=tool_kwargs,
                     )
-            elif function_name == "knowledge_search":
+            elif function_name == "file_search":
                 response_file_search_tool = next(
                     (t for t in ctx.response_tools if isinstance(t, OpenAIResponseInputToolFileSearch)),
                     None,
                 )
                 if response_file_search_tool:
-                    # Use vector_stores.search API instead of knowledge_search tool
+                    # Use vector_stores.search API instead of file_search tool
                     # to support filters and ranking_options
                     query = tool_kwargs.get("query", "")
-                    async with tracing.span("knowledge_search", {}):
+                    async with tracing.span("file_search", {}):
-                        result = await self._execute_knowledge_search_via_vector_store(
+                        result = await self._execute_file_search_via_vector_store(
                             query=query,
                             response_file_search_tool=response_file_search_tool,
                         )
@@ -341,7 +341,7 @@ class ToolExecutor:
                 output_index=output_index,
                 sequence_number=sequence_number,
             )
-        elif function_name == "knowledge_search":
+        elif function_name == "file_search":
             sequence_number += 1
             completion_event = OpenAIResponseObjectStreamResponseFileSearchCallCompleted(
                 item_id=item_id,
@@ -395,7 +395,7 @@ class ToolExecutor:
                 )
                 if has_error:
                     message.status = "failed"
-            elif function.name == "knowledge_search":
+            elif function.name == "file_search":
                 message = OpenAIResponseOutputMessageFileSearchToolCall(
                     id=item_id,
                     queries=[tool_kwargs.get("query", "")],

llama_stack/providers/inline/tool_runtime/file_search/file_search.py

@@ -226,9 +226,7 @@ class FileSearchToolRuntimeImpl(ToolGroupsProtocolPrivate, ToolRuntime, FileSear
 
         tokens = 0
         picked: list[InterleavedContentItem] = [
-            TextContentItem(
+            TextContentItem(text=f"file_search tool found {len(chunks)} chunks:\nBEGIN of file_search tool results.\n")
-                text=f"knowledge_search tool found {len(chunks)} chunks:\nBEGIN of knowledge_search tool results.\n"
-            )
         ]
         for i, chunk in enumerate(chunks):
             metadata = chunk.metadata
@@ -262,7 +260,7 @@ class FileSearchToolRuntimeImpl(ToolGroupsProtocolPrivate, ToolRuntime, FileSear
             text_content = query_config.chunk_template.format(index=i + 1, chunk=chunk, metadata=metadata_for_context)
             picked.append(TextContentItem(text=text_content))
 
-        picked.append(TextContentItem(text="END of knowledge_search tool results.\n"))
+        picked.append(TextContentItem(text="END of file_search tool results.\n"))
         picked.append(
             TextContentItem(
                 text=f'The above results were retrieved to help answer the user\'s query: "{interleaved_content_as_str(content)}". Use them as supporting information only in answering this query.\n',
@@ -292,7 +290,7 @@ class FileSearchToolRuntimeImpl(ToolGroupsProtocolPrivate, ToolRuntime, FileSear
                     description="Insert documents into memory",
                 ),
                 ToolDef(
-                    name="knowledge_search",
+                    name="file_search",
                     description="Search for information in a database.",
                     input_schema={
                         "type": "object",

llama_stack/ui/app/chat-playground/page.tsx

@@ -1167,7 +1167,10 @@ export default function ChatPlaygroundPage() {
 
     // find RAG toolgroups that have vector_db_ids configured
     const ragToolgroups = selectedAgentConfig.toolgroups.filter(toolgroup => {
-      if (typeof toolgroup === "object" && toolgroup.name?.includes("file_search")) {
+      if (
+        typeof toolgroup === "object" &&
+        toolgroup.name?.includes("file_search")
+      ) {
         return toolgroup.args && "vector_db_ids" in toolgroup.args;
       }
       return false;