alloy-runtime-cli 0.2.86__tar.gz → 0.2.87__tar.gz

{alloy_runtime_cli-0.2.86 → alloy_runtime_cli-0.2.87}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: alloy-runtime-cli
-Version: 0.2.86
+Version: 0.2.87
 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==0.2.86
-Requires-Dist: alloy-runtime-types==0.2.86
+Requires-Dist: alloy-runtime-sdk==0.2.87
+Requires-Dist: alloy-runtime-types==0.2.87
 Requires-Dist: trogon==0.6.0
 
 # alloy-runtime-cli
@@ -120,3 +120,11 @@ Use `alloy login` to save an API key locally, or set the environment variables w
 running on a VPS.
 
 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.
+
+## Hard-cut API migration
+
+CLI automation should use canonical `--model <provider>/<model>`, `--model-options`, and `--extra-headers`, and should parse the canonical `model` field from JSON output. Legacy fallback-chain and requested/actual model output assumptions are removed. See `docs/generate/hardcut_migration.md`.

{alloy_runtime_cli-0.2.86 → alloy_runtime_cli-0.2.87}/README.md

@@ -98,3 +98,11 @@ Use `alloy login` to save an API key locally, or set the environment variables w
 running on a VPS.
 
 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.
+
+## Hard-cut API migration
+
+CLI automation should use canonical `--model <provider>/<model>`, `--model-options`, and `--extra-headers`, and should parse the canonical `model` field from JSON output. Legacy fallback-chain and requested/actual model output assumptions are removed. See `docs/generate/hardcut_migration.md`.

{alloy_runtime_cli-0.2.86 → alloy_runtime_cli-0.2.87}/alloy_runtime_cli.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: alloy-runtime-cli
-Version: 0.2.86
+Version: 0.2.87
 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==0.2.86
-Requires-Dist: alloy-runtime-types==0.2.86
+Requires-Dist: alloy-runtime-sdk==0.2.87
+Requires-Dist: alloy-runtime-types==0.2.87
 Requires-Dist: trogon==0.6.0
 
 # alloy-runtime-cli
@@ -120,3 +120,11 @@ Use `alloy login` to save an API key locally, or set the environment variables w
 running on a VPS.
 
 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.
+
+## Hard-cut API migration
+
+CLI automation should use canonical `--model <provider>/<model>`, `--model-options`, and `--extra-headers`, and should parse the canonical `model` field from JSON output. Legacy fallback-chain and requested/actual model output assumptions are removed. See `docs/generate/hardcut_migration.md`.

{alloy_runtime_cli-0.2.86 → alloy_runtime_cli-0.2.87}/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==0.2.86
-alloy-runtime-types==0.2.86
+alloy-runtime-sdk==0.2.87
+alloy-runtime-types==0.2.87
 trogon==0.6.0

{alloy_runtime_cli-0.2.86 → alloy_runtime_cli-0.2.87}/cli/commands/agents/create/command.py

@@ -10,19 +10,21 @@ from pydantic import ValidationError
 
 from cli.commands.agents.create.presenter import present_agent_create_success
 from cli.commands.flag_utils import (
+    merge_raw_and_convenience_options,
+    parse_json_object_flag,
     parse_tags_required,
     require_exclusive,
-    validate_provider,
+    validate_reasoning_effort,
     validate_scope,
 )
+from cli.commands.shared_flags import REASONING_EFFORT
 from cli.infrastructure.command import async_command, authenticated_client
 from cli.infrastructure.file_content import FileContentError, resolve_content
 from alloy_runtime_types.dtos.agents import (
-    AgentModelConfig,
     AgentToolConfig,
     CreateAgentRequest,
 )
-from alloy_runtime_types.dtos.generation import GenerationParameters
+from alloy_runtime_types.dtos.generation import ReasoningEffort
 from alloy_runtime_types.enums.ownership_scope import OwnershipScope
 
 REQUEST_HEADERS = Annotated[
@@ -42,17 +44,11 @@ def agents_create_command(
         "--name",
         help="Agent name",
     ),
-    provider: str = typer.Option(
-        "...",
-        "-p",
-        "--provider",
-        help="Provider (openai, anthropic, etc.)",
-    ),
     model: str = typer.Option(
         "...",
         "-M",
         "--model",
-        help="Model name (gpt-4o, claude-3-sonnet, etc.)",
+        help="Canonical model (provider/model), e.g. openai/gpt-4o-mini",
     ),
     tags: str = typer.Option(
         "...",
@@ -82,6 +78,7 @@ def agents_create_command(
         "--max-tokens",
         help="Max tokens to generate",
     ),
+    reasoning_effort: str | None = REASONING_EFFORT,
     system_template: Optional[str] = typer.Option(
         None,
         "--system-template",
@@ -155,31 +152,26 @@ def agents_create_command(
         "--stop",
         help="Stop sequences (string or comma-separated for multiple)",
     ),
-    logprobs: Optional[bool] = typer.Option(
-        None,
-        "--logprobs/--no-logprobs",
-        help="Include log probabilities in response",
-    ),
-    top_logprobs: Optional[int] = typer.Option(
-        None,
-        "--top-logprobs",
-        help="Number of top log probabilities to return (1-20)",
-    ),
     user: Optional[str] = typer.Option(
         None,
         "--user",
         help="End-user ID for tracking and abuse monitoring",
     ),
+    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.",
+    ),
 ) -> None:
     """Create a new AI agent.
 
     Examples:
-        alloy agents create -n "My Agent" -p openai -M gpt-4o -t "agent.llm"
-        alloy agents create -n "Writer" -p anthropic -M claude-3-sonnet -t "agent.creative" -d "Creative writer" -T 0.9
-        alloy agents create -n "Org Agent" -p openai -M gpt-4o -t "agent.llm,team.backend" --scope organization
-        alloy agents create -n "With Tools" -p openai -M gpt-4o -t "agent" --tools '[{"tool_id": "web_search"}]'
-        alloy agents create -n "With Tools" -p openai -M gpt-4o -t "agent" --tools-file tools.json
-        alloy agents create -n "With Headers" -p openai -M gpt-4o -t "agent" -H 'anthropic-beta:context-1m-2025-08-07'
+        alloy agents create -n "My Agent" -M openai/gpt-4o -t "agent.llm"
+        alloy agents create -n "Writer" -M anthropic/claude-3-sonnet -t "agent.creative" -d "Creative writer" -T 0.9
+        alloy agents create -n "Org Agent" -M openai/gpt-4o -t "agent.llm,team.backend" --scope organization
+        alloy agents create -n "With Tools" -M openai/gpt-4o -t "agent" --tools '[{"tool_id": "web_search"}]'
+        alloy agents create -n "With Tools" -M openai/gpt-4o -t "agent" --tools-file tools.json
+        alloy agents create -n "With Headers" -M openai/gpt-4o -t "agent" -H 'anthropic-beta:context-1m-2025-08-07'
     """
     require_exclusive(("--tools", tools), ("--tools-file", tools_file))
     require_exclusive(
@@ -188,13 +180,13 @@ def agents_create_command(
 
     _execute_create_from_flags(
         name=name,
-        provider=provider,
         model=model,
         tags=tags,
         description=description,
         scope=scope,
         temperature=temperature,
         max_tokens=max_tokens,
+        reasoning_effort=reasoning_effort,
         system_template=system_template,
         tools=tools,
         tools_file=tools_file,
@@ -210,9 +202,11 @@ def agents_create_command(
         presence_penalty=presence_penalty,
         seed=seed,
         stop=stop,
-        logprobs=logprobs,
-        top_logprobs=top_logprobs,
         user=user,
+        raw_default_model_options=parse_json_object_flag(
+            default_model_options,
+            "--default-model-options",
+        ),
     )
 
 
@@ -376,15 +370,53 @@ def parse_stop_sequences(stop: Optional[str]) -> str | list[str] | None:
     return stop
 
 
+def _build_default_model_options(
+    *,
+    temperature: Optional[float],
+    max_tokens: Optional[int],
+    reasoning_effort: ReasoningEffort | None,
+    top_p: Optional[float],
+    top_k: Optional[int],
+    frequency_penalty: Optional[float],
+    presence_penalty: Optional[float],
+    seed: Optional[int],
+    stop: Optional[str],
+    user: Optional[str],
+) -> dict[str, Any] | None:
+    default_model_options: dict[str, Any] = {}
+    if temperature is not None:
+        default_model_options["temperature"] = temperature
+    if max_tokens is not None:
+        default_model_options["max_tokens"] = max_tokens
+    if reasoning_effort is not None:
+        default_model_options["reasoning_effort"] = reasoning_effort
+    if top_p is not None:
+        default_model_options["top_p"] = top_p
+    if top_k is not None:
+        default_model_options["top_k"] = top_k
+    if frequency_penalty is not None:
+        default_model_options["frequency_penalty"] = frequency_penalty
+    if presence_penalty is not None:
+        default_model_options["presence_penalty"] = presence_penalty
+    if seed is not None:
+        default_model_options["seed"] = seed
+    parsed_stop = parse_stop_sequences(stop)
+    if parsed_stop is not None:
+        default_model_options["stop"] = parsed_stop
+    if user is not None:
+        default_model_options["user"] = user
+    return default_model_options or None
+
+
 def _execute_create_from_flags(
     name: str,
-    provider: str,
     model: str,
     tags: str,
     description: Optional[str],
     scope: Optional[str],
     temperature: Optional[float],
     max_tokens: Optional[int],
+    reasoning_effort: str | None,
     system_template: Optional[str],
     tools: Optional[str],
     tools_file: Optional[Path],
@@ -400,54 +432,38 @@ def _execute_create_from_flags(
     presence_penalty: Optional[float],
     seed: Optional[int],
     stop: Optional[str],
-    logprobs: Optional[bool],
-    top_logprobs: Optional[int],
     user: Optional[str],
+    raw_default_model_options: dict[str, Any] | None = None,
 ) -> None:
     """Build and execute agent creation from command flags."""
-    provider_enum = validate_provider(provider)
-
     tag_list = parse_tags_required(tags)
 
-    model_config = AgentModelConfig(
-        provider_key=provider_enum,
-        provider_model_name=model,
-        preference_order=0,
-    )
-
     ownership_scope = OwnershipScope.ORGANIZATION
     if scope:
         ownership_scope = validate_scope(scope)
 
-    gen_params = None
-    if any(
-        [
-            temperature is not None,
-            max_tokens is not None,
-            top_p is not None,
-            top_k is not None,
-            frequency_penalty is not None,
-            presence_penalty is not None,
-            seed is not None,
-            stop is not None,
-            logprobs is not None,
-            top_logprobs is not None,
-            user is not None,
-        ]
-    ):
-        gen_params = GenerationParameters(
-            temperature=temperature,
-            max_tokens=max_tokens,
-            top_p=top_p,
-            top_k=top_k,
-            frequency_penalty=frequency_penalty,
-            presence_penalty=presence_penalty,
-            seed=seed,
-            stop=parse_stop_sequences(stop),
-            logprobs=logprobs,
-            top_logprobs=top_logprobs,
-            user=user,
-        )
+    validated_reasoning_effort = (
+        validate_reasoning_effort(reasoning_effort)
+        if reasoning_effort is not None
+        else None
+    )
+
+    convenience_default_model_options = _build_default_model_options(
+        temperature=temperature,
+        max_tokens=max_tokens,
+        reasoning_effort=validated_reasoning_effort,
+        top_p=top_p,
+        top_k=top_k,
+        frequency_penalty=frequency_penalty,
+        presence_penalty=presence_penalty,
+        seed=seed,
+        stop=stop,
+        user=user,
+    )
+    default_model_options = merge_raw_and_convenience_options(
+        raw_default_model_options,
+        convenience_default_model_options,
+    )
 
     tool_configs = parse_tools(tools, tools_file)
 
@@ -457,19 +473,19 @@ def _execute_create_from_flags(
 
     request = CreateAgentRequest(
         name=name,
-        models=[model_config],
+        model=model,
         tags=tag_list,
         scope=ownership_scope,
         description=description,
         system_instruction_template=system_template,
-        generation_params=gen_params,
+        default_model_options=default_model_options,
         tools=tool_configs,
         base_agent_id=base_agent,
         system_instruction_version_id=system_version,
         input_schema_id=input_schema,
         output_schema_id=output_schema,
         output_schema=parsed_output_schema,
-        default_request_headers=parsed_headers,
+        default_extra_headers=parsed_headers,
     )
 
     _execute_create(request)

{alloy_runtime_cli-0.2.86 → alloy_runtime_cli-0.2.87}/cli/commands/agents/create/fields.py

@@ -7,13 +7,14 @@ from cli.infrastructure.formatting.fields import (
 )
 from cli.infrastructure.renderers.entity_renderer import FieldConfig
 from alloy_runtime_types.dtos.agents import CreateAgentResponse
+from cli.commands.agents.tool_formatting import (
+    format_generation_params as _format_generation_params,
+)
 
 
 def format_models(response: CreateAgentResponse) -> str:
-    """Format models list for display."""
-    return ", ".join(
-        f"{m.provider_key.value}:{m.provider_model_name}" for m in response.models
-    )
+    """Format singular model for display."""
+    return response.model
 
 
 def format_tools(response: CreateAgentResponse) -> str:
@@ -31,19 +32,11 @@ def format_tags(response: CreateAgentResponse) -> str:
 
 
 def format_generation_params(response: CreateAgentResponse) -> str:
-    """Format generation parameters for display."""
-    if not response.generation_params:
-        return "(default)"
-
-    params: list[str] = []
-    if response.generation_params.temperature is not None:
-        params.append(f"temp={response.generation_params.temperature}")
-    if response.generation_params.max_tokens is not None:
-        params.append(f"max_tokens={response.generation_params.max_tokens}")
-    if response.generation_params.reasoning_effort is not None:
-        params.append(f"reasoning={response.generation_params.reasoning_effort}")
-
-    return ", ".join(params) if params else "(default)"
+    """Format complete generation defaults for display."""
+    return _format_generation_params(
+        model_options=response.default_model_options,
+        extra_headers=response.default_extra_headers,
+    )
 
 
 AGENT_FIELDS = [

{alloy_runtime_cli-0.2.86 → alloy_runtime_cli-0.2.87}/cli/commands/agents/create/presenter.py

@@ -2,7 +2,6 @@
 
 from cli.commands.agents.create.fields import (
     format_generation_params,
-    format_models,
     format_tags,
     format_tools,
 )
@@ -38,7 +37,7 @@ def present_agent_create_success(response: CreateAgentResponse) -> None:
         FieldConfig("id", "ID", lambda v: format_uuid(v, short=False)),
         FieldConfig("name", "Name"),
         FieldConfig("description", "Description", format_optional),
-        FieldConfig("models", "Models", lambda _: format_models(response)),
+        FieldConfig("model", "Model"),
         FieldConfig("tools", "Tools", lambda _: format_tools(response)),
         FieldConfig("tags", "Tags", lambda _: format_tags(response)),
         FieldConfig(

{alloy_runtime_cli-0.2.86 → alloy_runtime_cli-0.2.87}/cli/commands/agents/get/presenter.py

@@ -2,15 +2,14 @@
 
 from alloy_runtime_types.dtos.agents import GetAgentResponse
 
-from cli.commands.agents.tool_formatting import format_tool_details
+from cli.commands.agents.tool_formatting import (
+    format_generation_params,
+    format_tool_details,
+)
 from cli.infrastructure.output import OutputService
 from cli.infrastructure.renderers.entity_renderer import FieldConfig
 
 
-def _format_configured_count(values: list[object]) -> str:
-    return f"{len(values)} configured"
-
-
 def _format_tags(response: GetAgentResponse) -> str:
     if not response.tags:
         return "[]"
@@ -23,7 +22,16 @@ def present_agent_details(response: GetAgentResponse) -> None:
         FieldConfig("id", "ID", str),
         FieldConfig("name", "Name"),
         FieldConfig("description", "Description"),
-        FieldConfig("models", "Models", _format_configured_count),
+        FieldConfig("model", "Model"),
+        FieldConfig(
+            "default_model_options",
+            "Generation Params",
+            lambda _: format_generation_params(
+                model_options=response.default_model_options,
+                extra_headers=response.default_extra_headers,
+                empty="{}",
+            ),
+        ),
         FieldConfig("tools", "Tools", lambda _: format_tool_details(response.tools)),
         FieldConfig("id", "Tags", lambda _: _format_tags(response)),
     ]

{alloy_runtime_cli-0.2.86 → alloy_runtime_cli-0.2.87}/cli/commands/agents/list/presenter.py

@@ -14,11 +14,10 @@ def _format_created_at(value: datetime | None) -> str:
     return value.strftime("%Y-%m-%d")
 
 
-def _format_model_count(value: int | None) -> str:
+def _format_model(value: str | None) -> str:
     if value is None:
         return "(null)"
-    suffix = "model" if value == 1 else "models"
-    return f"{value} {suffix}"
+    return value
 
 
 def _format_description(value: str | None) -> str:
@@ -39,9 +38,9 @@ def present_agents_list(response: ListAgentsResponse) -> None:
             header_label="ID",
         ),
         ColumnConfig(
-            source_field="model_count",
-            header_label="Models",
-            formatter=_format_model_count,
+            source_field="model",
+            header_label="Model",
+            formatter=_format_model,
         ),
         ColumnConfig(
             source_field="created_at",

{alloy_runtime_cli-0.2.86 → alloy_runtime_cli-0.2.87}/cli/commands/agents/tool_formatting.py

@@ -1,5 +1,7 @@
 import json
 
+from typing import Any
+
 from alloy_runtime_types.dtos.agents import AgentToolResponse
 
 
@@ -23,3 +25,19 @@ def format_tool_details(tools: list[AgentToolResponse]) -> str:
             lines.append(f"tool_config_name={tool.tool_config_name}")
         parts.append(" ".join(lines))
     return "\n\n".join(parts)
+
+
+def format_generation_params(
+    *,
+    model_options: dict[str, Any] | None,
+    extra_headers: dict[str, str] | None,
+    empty: str = "(default)",
+) -> str:
+    payload: dict[str, Any] = {}
+    if model_options:
+        payload["model_options"] = model_options
+    if extra_headers:
+        payload["extra_headers"] = extra_headers
+    if not payload:
+        return empty
+    return json.dumps(payload, sort_keys=True)