alloy-runtime-cli 0.2.93__tar.gz → 0.2.95__tar.gz

{alloy_runtime_cli-0.2.93 → alloy_runtime_cli-0.2.95}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: alloy-runtime-cli
-Version: 0.2.93
+Version: 0.2.95
 Summary: Alloy Runtime CLI - Command-line interface for Alloy Runtime server
 Keywords: alloy,cli,python,terminal,workflow
 Classifier: Development Status :: 3 - Alpha
@@ -16,8 +16,8 @@ Requires-Dist: textual==8.2.3
 Requires-Dist: genson==1.3.0
 Requires-Dist: pyperclip==1.11.0
 Requires-Dist: prompt-toolkit==3.0.52
-Requires-Dist: alloy-runtime-sdk==0.2.93
-Requires-Dist: alloy-runtime-types==0.2.93
+Requires-Dist: alloy-runtime-sdk==0.2.95
+Requires-Dist: alloy-runtime-types==0.2.95
 Requires-Dist: trogon==0.6.0
 
 # alloy-runtime-cli
@@ -123,7 +123,7 @@ CLI release trigger: no functional changes.
 
 ## Model options and header display
 
-Agent create/update/get output displays complete `default_model_options` maps instead of compacting to a few common keys. Header defaults are displayed with redacted values because server read surfaces never return raw `default_extra_headers`. `alloy models list` shows the copyable canonical `<provider>/<model>` identity in the Model column.
+Agent create/update/get output displays complete `default_model_options` maps instead of compacting to a few common keys. On `agents update`, use `--clear-default-model-options` to clear persisted defaults to `null`, or `--default-model-options '{}'` to replace them with an empty object. Header defaults are displayed with redacted values because server read surfaces never return raw `default_extra_headers`. `alloy models list` shows the copyable canonical `<provider>/<model>` identity in the Model column.
 
 ## Hard-cut API migration
 

{alloy_runtime_cli-0.2.93 → alloy_runtime_cli-0.2.95}/README.md

@@ -101,7 +101,7 @@ CLI release trigger: no functional changes.
 
 ## Model options and header display
 
-Agent create/update/get output displays complete `default_model_options` maps instead of compacting to a few common keys. Header defaults are displayed with redacted values because server read surfaces never return raw `default_extra_headers`. `alloy models list` shows the copyable canonical `<provider>/<model>` identity in the Model column.
+Agent create/update/get output displays complete `default_model_options` maps instead of compacting to a few common keys. On `agents update`, use `--clear-default-model-options` to clear persisted defaults to `null`, or `--default-model-options '{}'` to replace them with an empty object. Header defaults are displayed with redacted values because server read surfaces never return raw `default_extra_headers`. `alloy models list` shows the copyable canonical `<provider>/<model>` identity in the Model column.
 
 ## Hard-cut API migration
 

{alloy_runtime_cli-0.2.93 → alloy_runtime_cli-0.2.95}/alloy_runtime_cli.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: alloy-runtime-cli
-Version: 0.2.93
+Version: 0.2.95
 Summary: Alloy Runtime CLI - Command-line interface for Alloy Runtime server
 Keywords: alloy,cli,python,terminal,workflow
 Classifier: Development Status :: 3 - Alpha
@@ -16,8 +16,8 @@ Requires-Dist: textual==8.2.3
 Requires-Dist: genson==1.3.0
 Requires-Dist: pyperclip==1.11.0
 Requires-Dist: prompt-toolkit==3.0.52
-Requires-Dist: alloy-runtime-sdk==0.2.93
-Requires-Dist: alloy-runtime-types==0.2.93
+Requires-Dist: alloy-runtime-sdk==0.2.95
+Requires-Dist: alloy-runtime-types==0.2.95
 Requires-Dist: trogon==0.6.0
 
 # alloy-runtime-cli
@@ -123,7 +123,7 @@ CLI release trigger: no functional changes.
 
 ## Model options and header display
 
-Agent create/update/get output displays complete `default_model_options` maps instead of compacting to a few common keys. Header defaults are displayed with redacted values because server read surfaces never return raw `default_extra_headers`. `alloy models list` shows the copyable canonical `<provider>/<model>` identity in the Model column.
+Agent create/update/get output displays complete `default_model_options` maps instead of compacting to a few common keys. On `agents update`, use `--clear-default-model-options` to clear persisted defaults to `null`, or `--default-model-options '{}'` to replace them with an empty object. Header defaults are displayed with redacted values because server read surfaces never return raw `default_extra_headers`. `alloy models list` shows the copyable canonical `<provider>/<model>` identity in the Model column.
 
 ## Hard-cut API migration
 

{alloy_runtime_cli-0.2.93 → alloy_runtime_cli-0.2.95}/alloy_runtime_cli.egg-info/requires.txt

@@ -4,6 +4,6 @@ textual==8.2.3
 genson==1.3.0
 pyperclip==1.11.0
 prompt-toolkit==3.0.52
-alloy-runtime-sdk==0.2.93
-alloy-runtime-types==0.2.93
+alloy-runtime-sdk==0.2.95
+alloy-runtime-types==0.2.95
 trogon==0.6.0

{alloy_runtime_cli-0.2.93 → alloy_runtime_cli-0.2.95}/cli/commands/agents/update/command.py

@@ -201,7 +201,12 @@ def agents_update_command(
     default_model_options: Optional[str] = typer.Option(
         None,
         "--default-model-options",
-        help="Raw default model options as JSON object or @file.json. Convenience flags override duplicate keys.",
+        help="Raw default model options as JSON object or @file.json. Convenience flags override duplicate keys. Use '{}' to replace with an empty object.",
+    ),
+    clear_default_model_options: bool = typer.Option(
+        False,
+        "--clear-default-model-options",
+        help="Clear all default model options from the agent.",
     ),
 ) -> None:
     """Update an existing agent.
@@ -222,12 +227,30 @@ def agents_update_command(
         alloy agents update my-agent --top-p 0.9 --frequency-penalty 0.5
         alloy agents update my-agent --output-schema-def '{"type": "object"}'
         alloy agents update my-agent -H 'anthropic-beta:context-1m-2025-08-07'
+        alloy agents update my-agent --clear-default-model-options
     """
     require_exclusive(
         ("--tools", tools),
         ("--tools-file", tools_file),
         ("--clear-tools", clear_tools),
-        ("--default-model-options", default_model_options),
+    )
+
+    has_default_model_options_update = (
+        default_model_options is not None
+        or temperature is not None
+        or max_tokens is not None
+        or top_p is not None
+        or top_k is not None
+        or frequency_penalty is not None
+        or presence_penalty is not None
+        or reasoning_effort is not None
+        or seed is not None
+        or stop is not None
+        or user is not None
+    )
+    require_exclusive(
+        ("--clear-default-model-options", clear_default_model_options),
+        ("default model option update flags", has_default_model_options_update),
     )
 
     require_exclusive(
@@ -298,6 +321,8 @@ def agents_update_command(
         ("--tools", tools),
         ("--tools-file", tools_file),
         ("--clear-tools", clear_tools),
+        ("--default-model-options", default_model_options),
+        ("--clear-default-model-options", clear_default_model_options),
     )
 
     validated_reasoning_effort = (
@@ -333,6 +358,7 @@ def agents_update_command(
         tools=tools,
         tools_file=tools_file,
         clear_tools=clear_tools,
+        clear_default_model_options=clear_default_model_options,
         raw_default_model_options=parse_json_object_flag(
             default_model_options,
             "--default-model-options",
@@ -368,6 +394,7 @@ async def _execute_update(
     tools: Optional[str],
     tools_file: Optional[Path],
     clear_tools: bool,
+    clear_default_model_options: bool,
     raw_default_model_options: dict[str, Any] | None = None,
 ) -> None:
     """Execute the agent update operation."""
@@ -423,8 +450,10 @@ async def _execute_update(
         raw_default_model_options,
         convenience_default_model_options,
     )
-    if default_model_options is not None:
-        request_data["default_model_options"] = default_model_options
+    if clear_default_model_options:
+        request_data["default_model_options"] = None
+    elif raw_default_model_options is not None or default_model_options is not None:
+        request_data["default_model_options"] = default_model_options or {}
 
     if clear_tools:
         request_data["tools"] = []

{alloy_runtime_cli-0.2.93 → alloy_runtime_cli-0.2.95}/cli/commands/generate/text/command.py

@@ -238,11 +238,6 @@ def generate_text_command(
         "--webhook-url",
         help="Webhook URL for completion notification",
     ),
-    tools: Optional[str] = typer.Option(
-        None,
-        "--tools",
-        help="Tool definitions as JSON or @file.json (direct model mode)",
-    ),
     extra_headers: Optional[str] = typer.Option(
         None,
         "--extra-headers",
@@ -379,13 +374,6 @@ def generate_text_command(
             "Async execution does not support multi-agent comparison."
         )
 
-    # Validate --tools can only be used with direct model mode (not agents)
-    if tools and (agent or agents):
-        raise typer.BadParameter(
-            "--tools cannot be used with --agent or --agents. "
-            "For agents, configure tools in the agent's settings."
-        )
-
     # Validate mutually exclusive options
     validate_arguments(
         message=message,
@@ -427,7 +415,6 @@ def generate_text_command(
 
     # Parse JSON options
     parsed_user_metadata = parse_json_or_file(user_metadata, "--user-metadata")
-    parsed_tools = parse_json_or_file(tools, "--tools")
     parsed_extra_headers = parse_json_or_file(
         extra_headers,
         "--extra-headers",
@@ -485,7 +472,6 @@ def generate_text_command(
         regenerate=regenerate,
         user_metadata=parsed_user_metadata,
         webhook_url=webhook_url,
-        tools=parsed_tools,
         parent_execution_id=parent_execution_id,
         execution_purpose=execution_purpose,
     )
@@ -831,7 +817,6 @@ async def _execute_generate_text(
     regenerate: bool = False,
     user_metadata: dict[str, Any] | None = None,
     webhook_url: str | None = None,
-    tools: list[Any] | None = None,
     parent_execution_id: UUID | None = None,
     execution_purpose: str | None = None,
     show_full_trace: bool = False,
@@ -881,6 +866,8 @@ async def _execute_generate_text(
             schema_definition=schema_def,
         )
 
+    request_mode = "async" if run_async else "stream" if stream else "sync"
+
     # Multi-agent comparison mode
     if agents:
         await _execute_multi_agent(
@@ -895,6 +882,7 @@ async def _execute_generate_text(
             extra_headers=extra_headers,
             tag_list=tag_list,
             stream=stream,
+            request_mode=request_mode,
             show_thinking=show_thinking,
             show_full_trace=show_full_trace,
             console=console,
@@ -904,32 +892,35 @@ async def _execute_generate_text(
         )
         return
 
-    request = GenerateTextRequest(
-        agent_id=agent,
-        model=model,
-        prompt=message,
-        prompt_template_id=message_template,
-        prompt_variables=message_variables,
-        system=system_instruction,
-        system_template_id=system_template,
-        system_variables=system_variables,
-        chat_session_id=session_id,
-        tags=tag_list,
-        stream=stream,
-        output_schema=output_schema,
-        model_options=model_options,
-        extra_headers=extra_headers,
-        show_full_trace=show_full_trace,
-        external_id=external_id,
-        billing_project_id=billing_project_id,
-        max_history_messages=max_history_messages,
-        regenerate=regenerate,
-        user_metadata=user_metadata,
-        webhook_url=webhook_url,
-        tools=tools,
-        parent_execution_id=parent_execution_id,
-        execution_purpose=execution_purpose,
-    )
+    request_data: dict[str, Any] = {
+        "agent_id": agent,
+        "model": model,
+        "prompt": message,
+        "prompt_template_id": message_template,
+        "prompt_variables": message_variables,
+        "system": system_instruction,
+        "system_template_id": system_template,
+        "system_variables": system_variables,
+        "chat_session_id": session_id,
+        "tags": tag_list,
+        "output_schema": output_schema,
+        "model_options": model_options,
+        "extra_headers": extra_headers,
+        "show_full_trace": show_full_trace,
+        "external_id": external_id,
+        "billing_project_id": billing_project_id,
+        "max_history_messages": max_history_messages,
+        "regenerate": regenerate,
+        "user_metadata": user_metadata,
+        "webhook_url": webhook_url,
+        "parent_execution_id": parent_execution_id,
+        "execution_purpose": execution_purpose,
+    }
+    if "mode" in GenerateTextRequest.model_fields:
+        request_data["mode"] = request_mode
+    else:
+        request_data["stream"] = stream
+    request = GenerateTextRequest(**request_data)
 
     # Handle async execution mode
     if run_async:
@@ -1163,6 +1154,7 @@ async def _execute_multi_agent(
     extra_headers: dict[str, str] | None,
     tag_list: list[str],
     stream: bool,
+    request_mode: str,
     show_thinking: bool,
     console: Console,
     _output: OutputService,
@@ -1171,23 +1163,26 @@ async def _execute_multi_agent(
     show_full_trace: bool = False,
 ) -> None:
     async with authenticated_client() as (_config, client):
-        requests = [
-            GenerateTextRequest(
-                agent_id=agent_id,
-                prompt=message,
-                prompt_template_id=message_template,
-                prompt_variables=message_variables,
-                system=system_instruction,
-                system_template_id=system_template,
-                system_variables=system_variables,
-                tags=tag_list,
-                stream=stream,
-                model_options=model_options,
-                extra_headers=extra_headers,
-                show_full_trace=show_full_trace,
-            )
-            for agent_id in agents
-        ]
+        requests: list[GenerateTextRequest] = []
+        for agent_id in agents:
+            request_data: dict[str, Any] = {
+                "agent_id": agent_id,
+                "prompt": message,
+                "prompt_template_id": message_template,
+                "prompt_variables": message_variables,
+                "system": system_instruction,
+                "system_template_id": system_template,
+                "system_variables": system_variables,
+                "tags": tag_list,
+                "model_options": model_options,
+                "extra_headers": extra_headers,
+                "show_full_trace": show_full_trace,
+            }
+            if "mode" in GenerateTextRequest.model_fields:
+                request_data["mode"] = request_mode
+            else:
+                request_data["stream"] = stream
+            requests.append(GenerateTextRequest(**request_data))
 
         if stream:
             renderer = ConcurrentRenderer(

{alloy_runtime_cli-0.2.93 → alloy_runtime_cli-0.2.95}/cli/tui/chat/commands/regenerate.py

@@ -1,6 +1,7 @@
 """Command for regenerating the last AI response."""
 
 from dataclasses import dataclass
+from typing import Any
 from uuid import UUID
 
 from alloy_runtime_types.dtos.generation import GenerateTextRequest
@@ -60,23 +61,31 @@ class RegenerateCommand:
         try:
             # Build request based on mode
             if self.agent_id is not None:
-                request = GenerateTextRequest(
-                    agent_id=self.agent_id,
-                    chat_session_id=self.session_id,
-                    regenerate=True,
-                    stream=False,
-                    tags=[],
-                )
+                request_data: dict[str, Any] = {
+                    "agent_id": self.agent_id,
+                    "chat_session_id": self.session_id,
+                    "regenerate": True,
+                    "tags": [],
+                }
+                if "mode" in GenerateTextRequest.model_fields:
+                    request_data["mode"] = "sync"
+                else:
+                    request_data["stream"] = False
+                request = GenerateTextRequest(**request_data)
             elif self.provider_key and self.provider_model_name:
-                request = GenerateTextRequest(
-                    model=f"{self.provider_key}/{self.provider_model_name}",
-                    chat_session_id=self.session_id,
-                    regenerate=True,
-                    stream=False,
-                    tags=[],
+                request_data = {
+                    "model": f"{self.provider_key}/{self.provider_model_name}",
+                    "chat_session_id": self.session_id,
+                    "regenerate": True,
+                    "tags": [],
                     # Only include system instruction for model mode
-                    system=self.system_instruction,
-                )
+                    "system": self.system_instruction,
+                }
+                if "mode" in GenerateTextRequest.model_fields:
+                    request_data["mode"] = "sync"
+                else:
+                    request_data["stream"] = False
+                request = GenerateTextRequest(**request_data)
             else:
                 self.store.finish_generation()
                 return CommandResult[None](

{alloy_runtime_cli-0.2.93 → alloy_runtime_cli-0.2.95}/cli/tui/chat/commands/send_message.py

@@ -2,6 +2,7 @@
 
 from collections.abc import Callable
 from dataclasses import dataclass
+from typing import Any
 from uuid import UUID
 
 from alloy_runtime_types.dtos.generation import GenerateTextRequest
@@ -121,23 +122,31 @@ class SendMessageCommand:
         try:
             # Build request based on mode
             if self.agent_id is not None:
-                request = GenerateTextRequest(
-                    agent_id=self.agent_id,
-                    prompt=message_to_send,  # Use resolved message
-                    stream=False,
-                    tags=[],
-                    chat_session_id=self.session_id,
-                )
+                request_data: dict[str, Any] = {
+                    "agent_id": self.agent_id,
+                    "prompt": message_to_send,  # Use resolved message
+                    "tags": [],
+                    "chat_session_id": self.session_id,
+                }
+                if "mode" in GenerateTextRequest.model_fields:
+                    request_data["mode"] = "sync"
+                else:
+                    request_data["stream"] = False
+                request = GenerateTextRequest(**request_data)
             elif self.provider_key and self.provider_model_name:
-                request = GenerateTextRequest(
-                    model=f"{self.provider_key}/{self.provider_model_name}",
-                    prompt=message_to_send,  # Use resolved message
-                    stream=False,
-                    tags=[],
-                    chat_session_id=self.session_id,
+                request_data = {
+                    "model": f"{self.provider_key}/{self.provider_model_name}",
+                    "prompt": message_to_send,  # Use resolved message
+                    "tags": [],
+                    "chat_session_id": self.session_id,
                     # Only include system instruction for model mode
-                    system=self.system_instruction,
-                )
+                    "system": self.system_instruction,
+                }
+                if "mode" in GenerateTextRequest.model_fields:
+                    request_data["mode"] = "sync"
+                else:
+                    request_data["stream"] = False
+                request = GenerateTextRequest(**request_data)
             else:
                 self.store.finish_generation()
                 return CommandResult[None](

{alloy_runtime_cli-0.2.93 → alloy_runtime_cli-0.2.95}/cli/tui/chat/services/name_generator.py

@@ -4,6 +4,8 @@ Automatically generates meaningful session names using the current session model
 when it is cheap enough for a lightweight naming request.
 """
 
+from typing import Any
+
 from alloy_runtime_sdk.api_client.client import ApiClient
 from alloy_runtime_types.dtos.generation import GenerateTextRequest
 from alloy_runtime_sdk.logging.config import get_logger
@@ -129,12 +131,16 @@ class SessionNameGenerator:
                 model=session_model_name,
             )
 
-            request = GenerateTextRequest(
-                model=f"{session_provider_key}/{session_model_name}",
-                prompt=prompt,
-                stream=False,
-                tags=["system.session_naming"],
-            )
+            request_data: dict[str, Any] = {
+                "model": f"{session_provider_key}/{session_model_name}",
+                "prompt": prompt,
+                "tags": ["system.session_naming"],
+            }
+            if "mode" in GenerateTextRequest.model_fields:
+                request_data["mode"] = "sync"
+            else:
+                request_data["stream"] = False
+            request = GenerateTextRequest(**request_data)
 
             response = await client.generate_text_sync(request)
             logger.info(

{alloy_runtime_cli-0.2.93 → alloy_runtime_cli-0.2.95}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "alloy-runtime-cli"
-version = "0.2.93"
+version = "0.2.95"
 description = "Alloy Runtime CLI - Command-line interface for Alloy Runtime server"
 readme = "README.md"
 requires-python = ">=3.13"
@@ -23,8 +23,8 @@ dependencies = [
   "genson==1.3.0",
   "pyperclip==1.11.0",
   "prompt-toolkit==3.0.52",
-  "alloy-runtime-sdk==0.2.93",
-  "alloy-runtime-types==0.2.93",
+  "alloy-runtime-sdk==0.2.95",
+  "alloy-runtime-types==0.2.95",
   "trogon==0.6.0",
 ]