feat(tools)!: substantial clean up of "Tool" related datatypes (#3627)

This is a sweeping change to clean up some gunk around our "Tool"
definitions.

First, we had two types `Tool` and `ToolDef`. The first of these was a
"Resource" type for the registry but we had stopped registering tools
inside the Registry long back (and only registered ToolGroups.) The
latter was for specifying tools for the Agents API. This PR removes the
former and adds an optional `toolgroup_id` field to the latter.

Secondly, as pointed out by @bbrowning in
https://github.com/llamastack/llama-stack/pull/3003#issuecomment-3245270132,
we were doing a lossy conversion from a full JSON schema from the MCP
tool specification into our ToolDefinition to send it to the model.
There is no necessity to do this -- we ourselves aren't doing any
execution at all but merely passing it to the chat completions API which
supports this. By doing this (and by doing it poorly), we encountered
limitations like not supporting array items, or not resolving $refs,
etc.

To fix this, we replaced the `parameters` field by `{ input_schema,
output_schema }` which can be full blown JSON schemas.

Finally, there were some types in our llama-related chat format
conversion which needed some cleanup. We are taking this opportunity to
clean those up.

This PR is a substantial breaking change to the API. However, given our
window for introducing breaking changes, this suits us just fine. I will
be landing a concurrent `llama-stack-client` change as well since API
shapes are changing.

llama_stack/providers/remote/inference/vllm/vllm.py

@@ -89,8 +89,7 @@ def _convert_to_vllm_tool_calls_in_response(
         ToolCall(
             call_id=call.id,
             tool_name=call.function.name,
-            arguments=json.loads(call.function.arguments),
-            arguments_json=call.function.arguments,
+            arguments=call.function.arguments,
         )
         for call in tool_calls
     ]
@@ -100,18 +99,6 @@ def _convert_to_vllm_tools_in_request(tools: list[ToolDefinition]) -> list[dict]
     compat_tools = []
 
     for tool in tools:
-        properties = {}
-        compat_required = []
-        if tool.parameters:
-            for tool_key, tool_param in tool.parameters.items():
-                properties[tool_key] = {"type": tool_param.param_type}
-                if tool_param.description:
-                    properties[tool_key]["description"] = tool_param.description
-                if tool_param.default:
-                    properties[tool_key]["default"] = tool_param.default
-                if tool_param.required:
-                    compat_required.append(tool_key)
-
         # The tool.tool_name can be a str or a BuiltinTool enum. If
         # it's the latter, convert to a string.
         tool_name = tool.tool_name
@@ -123,10 +110,11 @@ def _convert_to_vllm_tools_in_request(tools: list[ToolDefinition]) -> list[dict]
             "function": {
                 "name": tool_name,
                 "description": tool.description,
-                "parameters": {
+                "parameters": tool.input_schema
+                or {
                     "type": "object",
-                    "properties": properties,
-                    "required": compat_required,
+                    "properties": {},
+                    "required": [],
                 },
             },
         }
@@ -161,7 +149,6 @@ def _process_vllm_chat_completion_end_of_stream(
     for _index, tool_call_buf in sorted(tool_call_bufs.items()):
         args_str = tool_call_buf.arguments or "{}"
         try:
-            args = json.loads(args_str)
             chunks.append(
                 ChatCompletionResponseStreamChunk(
                     event=ChatCompletionResponseEvent(
@@ -170,8 +157,7 @@ def _process_vllm_chat_completion_end_of_stream(
                             tool_call=ToolCall(
                                 call_id=tool_call_buf.call_id,
                                 tool_name=tool_call_buf.tool_name,
-                                arguments=args,
-                                arguments_json=args_str,
+                                arguments=args_str,
                             ),
                             parse_status=ToolCallParseStatus.succeeded,
                         ),

llama_stack/providers/remote/tool_runtime/bing_search/bing_search.py

@@ -15,7 +15,6 @@ from llama_stack.apis.tools import (
     ToolDef,
     ToolGroup,
     ToolInvocationResult,
-    ToolParameter,
     ToolRuntime,
 )
 from llama_stack.core.request_headers import NeedsRequestProviderData
@@ -57,13 +56,16 @@ class BingSearchToolRuntimeImpl(ToolGroupsProtocolPrivate, ToolRuntime, NeedsReq
                 ToolDef(
                     name="web_search",
                     description="Search the web using Bing Search API",
-                    parameters=[
-                        ToolParameter(
-                            name="query",
-                            description="The query to search for",
-                            parameter_type="string",
-                        )
-                    ],
+                    input_schema={
+                        "type": "object",
+                        "properties": {
+                            "query": {
+                                "type": "string",
+                                "description": "The query to search for",
+                            }
+                        },
+                        "required": ["query"],
+                    },
                 )
             ]
         )

llama_stack/providers/remote/tool_runtime/brave_search/brave_search.py

@@ -14,7 +14,6 @@ from llama_stack.apis.tools import (
     ToolDef,
     ToolGroup,
     ToolInvocationResult,
-    ToolParameter,
     ToolRuntime,
 )
 from llama_stack.core.request_headers import NeedsRequestProviderData
@@ -56,13 +55,16 @@ class BraveSearchToolRuntimeImpl(ToolGroupsProtocolPrivate, ToolRuntime, NeedsRe
                 ToolDef(
                     name="web_search",
                     description="Search the web for information",
-                    parameters=[
-                        ToolParameter(
-                            name="query",
-                            description="The query to search for",
-                            parameter_type="string",
-                        )
-                    ],
+                    input_schema={
+                        "type": "object",
+                        "properties": {
+                            "query": {
+                                "type": "string",
+                                "description": "The query to search for",
+                            }
+                        },
+                        "required": ["query"],
+                    },
                     built_in_type=BuiltinTool.brave_search,
                 )
             ]

llama_stack/providers/remote/tool_runtime/tavily_search/tavily_search.py

@@ -15,7 +15,6 @@ from llama_stack.apis.tools import (
     ToolDef,
     ToolGroup,
     ToolInvocationResult,
-    ToolParameter,
     ToolRuntime,
 )
 from llama_stack.core.request_headers import NeedsRequestProviderData
@@ -56,13 +55,16 @@ class TavilySearchToolRuntimeImpl(ToolGroupsProtocolPrivate, ToolRuntime, NeedsR
                 ToolDef(
                     name="web_search",
                     description="Search the web for information",
-                    parameters=[
-                        ToolParameter(
-                            name="query",
-                            description="The query to search for",
-                            parameter_type="string",
-                        )
-                    ],
+                    input_schema={
+                        "type": "object",
+                        "properties": {
+                            "query": {
+                                "type": "string",
+                                "description": "The query to search for",
+                            }
+                        },
+                        "required": ["query"],
+                    },
                 )
             ]
         )

llama_stack/providers/remote/tool_runtime/wolfram_alpha/wolfram_alpha.py

@@ -15,7 +15,6 @@ from llama_stack.apis.tools import (
     ToolDef,
     ToolGroup,
     ToolInvocationResult,
-    ToolParameter,
     ToolRuntime,
 )
 from llama_stack.core.request_headers import NeedsRequestProviderData
@@ -57,13 +56,16 @@ class WolframAlphaToolRuntimeImpl(ToolGroupsProtocolPrivate, ToolRuntime, NeedsR
                 ToolDef(
                     name="wolfram_alpha",
                     description="Query WolframAlpha for computational knowledge",
-                    parameters=[
-                        ToolParameter(
-                            name="query",
-                            description="The query to compute",
-                            parameter_type="string",
-                        )
-                    ],
+                    input_schema={
+                        "type": "object",
+                        "properties": {
+                            "query": {
+                                "type": "string",
+                                "description": "The query to compute",
+                            }
+                        },
+                        "required": ["query"],
+                    },
                 )
             ]
         )