alloy-runtime-cli 0.2.71__tar.gz → 0.2.73__tar.gz

{alloy_runtime_cli-0.2.71 → alloy_runtime_cli-0.2.73}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: alloy-runtime-cli
-Version: 0.2.71
+Version: 0.2.73
 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==0.2.71
-Requires-Dist: alloy-runtime-types==0.2.71
+Requires-Dist: alloy-runtime-sdk==0.2.73
+Requires-Dist: alloy-runtime-types==0.2.73
 Requires-Dist: trogon==0.6.0
 
 # alloy-runtime-cli
@@ -77,7 +77,7 @@ alloy executions get <execution_id> --watch
 alloy executions get <execution_id> --full-output
 ```
 
-`alloy executions get` uses a dedicated transcript view for human output: one metadata header followed by chronological reasoning, tool-call, tool-output, and assistant blocks. Tool inputs are always complete; tool outputs are compact by default, and `--full-output` expands them. `--watch` appends only newly observed activity blocks while the execution is still running and waits indefinitely unless you pass `--timeout <seconds>`.
+`alloy executions get` uses a dedicated transcript view for human output: one metadata header followed by chronological reasoning, grouped tool-call/result, and assistant blocks. Tool inputs are always complete; tool outputs are compact by default, and `--full-output` expands them. `--watch` repaints a live grouped snapshot while the execution is still running and waits indefinitely unless you pass `--timeout <seconds>`.
 
 `alloy ping` is the authenticated connectivity check. It confirms who you are, which host you reached, and which server build answered the request.
 

{alloy_runtime_cli-0.2.71 → alloy_runtime_cli-0.2.73}/README.md

@@ -55,7 +55,7 @@ alloy executions get <execution_id> --watch
 alloy executions get <execution_id> --full-output
 ```
 
-`alloy executions get` uses a dedicated transcript view for human output: one metadata header followed by chronological reasoning, tool-call, tool-output, and assistant blocks. Tool inputs are always complete; tool outputs are compact by default, and `--full-output` expands them. `--watch` appends only newly observed activity blocks while the execution is still running and waits indefinitely unless you pass `--timeout <seconds>`.
+`alloy executions get` uses a dedicated transcript view for human output: one metadata header followed by chronological reasoning, grouped tool-call/result, and assistant blocks. Tool inputs are always complete; tool outputs are compact by default, and `--full-output` expands them. `--watch` repaints a live grouped snapshot while the execution is still running and waits indefinitely unless you pass `--timeout <seconds>`.
 
 `alloy ping` is the authenticated connectivity check. It confirms who you are, which host you reached, and which server build answered the request.
 

{alloy_runtime_cli-0.2.71 → alloy_runtime_cli-0.2.73}/alloy_runtime_cli.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: alloy-runtime-cli
-Version: 0.2.71
+Version: 0.2.73
 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==0.2.71
-Requires-Dist: alloy-runtime-types==0.2.71
+Requires-Dist: alloy-runtime-sdk==0.2.73
+Requires-Dist: alloy-runtime-types==0.2.73
 Requires-Dist: trogon==0.6.0
 
 # alloy-runtime-cli
@@ -77,7 +77,7 @@ alloy executions get <execution_id> --watch
 alloy executions get <execution_id> --full-output
 ```
 
-`alloy executions get` uses a dedicated transcript view for human output: one metadata header followed by chronological reasoning, tool-call, tool-output, and assistant blocks. Tool inputs are always complete; tool outputs are compact by default, and `--full-output` expands them. `--watch` appends only newly observed activity blocks while the execution is still running and waits indefinitely unless you pass `--timeout <seconds>`.
+`alloy executions get` uses a dedicated transcript view for human output: one metadata header followed by chronological reasoning, grouped tool-call/result, and assistant blocks. Tool inputs are always complete; tool outputs are compact by default, and `--full-output` expands them. `--watch` repaints a live grouped snapshot while the execution is still running and waits indefinitely unless you pass `--timeout <seconds>`.
 
 `alloy ping` is the authenticated connectivity check. It confirms who you are, which host you reached, and which server build answered the request.
 

{alloy_runtime_cli-0.2.71 → alloy_runtime_cli-0.2.73}/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==0.2.71
-alloy-runtime-types==0.2.71
+alloy-runtime-sdk==0.2.73
+alloy-runtime-types==0.2.73
 trogon==0.6.0

{alloy_runtime_cli-0.2.71 → alloy_runtime_cli-0.2.73}/cli/commands/executions/get/command.py

@@ -4,7 +4,10 @@ from uuid import UUID
 
 import typer
 
-from cli.commands.executions.get.presenter import present_execution
+from cli.commands.executions.get.presenter import (
+    present_execution,
+    watch_execution_live,
+)
 from cli.commands.flag_utils import validate_uuid
 from cli.infrastructure.command import async_command, authenticated_client
 from cli.infrastructure.output import OutputFormat, OutputService
@@ -52,9 +55,6 @@ async def _execute_get(
 
         output = OutputService.get()
         start_time = time.monotonic() if timeout is not None else None
-        metadata_rendered = False
-        rendered_count = 0
-        output.progress(f"Watching execution {execution_id}...")
 
         if output.format == OutputFormat.JSON:
             last_response = await client.get_execution(execution_id)
@@ -70,22 +70,12 @@ async def _execute_get(
             output.raw(last_response)
             return
 
-        while True:
-            response = await client.get_execution(execution_id)
-            rendered_count = present_execution(
-                response,
-                include_metadata=not metadata_rendered,
-                include_appendices=response.is_terminal,
-                start_index=rendered_count,
-                full_output=full_output,
-            )
-            metadata_rendered = True
-            if response.is_terminal:
-                return
-            if _watch_timed_out(start_time=start_time, timeout=timeout):
-                output.progress(f"Watch timed out after {timeout:.0f}s.")
-                return
-            await asyncio.sleep(_POLL_INTERVAL_SECONDS)
+        await watch_execution_live(
+            client=client,
+            execution_id=execution_id,
+            timeout=timeout,
+            full_output=full_output,
+        )
 
 
 def _watch_timed_out(start_time: float | None, timeout: float | None) -> bool:

alloy_runtime_cli-0.2.73/cli/commands/executions/get/presenter.py

@@ -0,0 +1,95 @@
+import asyncio
+import time
+from typing import Any
+from uuid import UUID
+
+from alloy_runtime_types.dtos.executions import GetExecutionResponse
+from cli.commands.executions.get.transcript_renderer import (
+    render_execution_snapshot,
+)
+from cli.infrastructure.output import OutputFormat, OutputService
+from rich.live import Live
+from rich.text import Text
+
+_LIVE_POLL_INTERVAL_SECONDS = 3.0
+
+
+def present_execution(
+    response: GetExecutionResponse,
+    *,
+    include_metadata: bool = True,
+    include_appendices: bool | None = None,
+    start_index: int = 0,
+    full_output: bool = False,
+) -> int:
+    output = OutputService.get()
+    if output.format == OutputFormat.JSON:
+        output.raw(response)
+        return len(response.activity)
+
+    should_render_appendices = (
+        include_metadata if include_appendices is None else include_appendices
+    )
+    rendered = render_execution_snapshot(
+        response,
+        include_metadata=include_metadata,
+        include_appendices=should_render_appendices,
+        start_index=start_index,
+        full_output=full_output,
+    )
+    if rendered:
+        output.text(rendered)
+
+    return len(response.activity)
+
+
+async def watch_execution_live(
+    *,
+    client: Any,
+    execution_id: UUID,
+    timeout: float | None,
+    full_output: bool,
+) -> None:
+    output = OutputService.get()
+    start_time = time.monotonic() if timeout is not None else None
+    response = await client.get_execution(execution_id)
+
+    with Live(
+        _to_live_renderable(response, full_output=full_output),
+        console=output.console,
+        auto_refresh=False,
+        refresh_per_second=4,
+        vertical_overflow="ellipsis",
+        transient=False,
+        redirect_stdout=False,
+        redirect_stderr=False,
+    ) as live:
+        live.refresh()
+        while not response.is_terminal:
+            if _watch_timed_out(start_time=start_time, timeout=timeout):
+                output.progress(f"Watch timed out after {timeout:.0f}s.")
+                return
+            await asyncio.sleep(_LIVE_POLL_INTERVAL_SECONDS)
+            response = await client.get_execution(execution_id)
+            live.update(
+                _to_live_renderable(response, full_output=full_output), refresh=True
+            )
+
+
+def _to_live_renderable(response: GetExecutionResponse, *, full_output: bool) -> Text:
+    return Text.from_markup(
+        render_execution_snapshot(
+            response,
+            include_metadata=True,
+            include_appendices=response.is_terminal,
+            full_output=full_output,
+        )
+    )
+
+
+def _watch_timed_out(start_time: float | None, timeout: float | None) -> bool:
+    if timeout is None:
+        return False
+    if start_time is None:
+        raise RuntimeError("watch timeout requires a start time")
+    return time.monotonic() - start_time >= timeout

{alloy_runtime_cli-0.2.71 → alloy_runtime_cli-0.2.73}/cli/commands/executions/get/transcript_renderer.py

@@ -2,7 +2,7 @@ import json
 import textwrap
 from dataclasses import dataclass
 from datetime import datetime
-from typing import Any, cast
+from typing import Any, TypeGuard, cast
 
 from alloy_runtime_types.dtos.executions import (
     AssistantTextActivity,
@@ -10,6 +10,7 @@ from alloy_runtime_types.dtos.executions import (
     GetExecutionResponse,
     ReasoningActivity,
     ToolCallActivity,
+    ToolErrorActivity,
     ToolOutputActivity,
 )
 from rich.markup import escape
@@ -71,11 +72,57 @@ def render_activity_blocks(
     width: int = 80,
 ) -> str:
     blocks: list[str] = []
-    for activity in activities[start_index:]:
+    activity_slice = activities[start_index:]
+    grouped_results = _group_tool_results(activity_slice)
+    rendered_result_indexes: set[int] = set()
+
+    for activity in activity_slice:
+        if isinstance(activity, ToolCallActivity):
+            results = grouped_results.get(activity.tool_call_id or "", [])
+            rendered_result_indexes.update(result.index for result in results)
+            blocks.append(
+                _render_tool_group(
+                    activity,
+                    results,
+                    full_output=full_output,
+                    width=width,
+                )
+            )
+            continue
+        if _is_tool_result(activity) and activity.index in rendered_result_indexes:
+            continue
         blocks.append(_render_activity(activity, full_output=full_output, width=width))
     return "\n\n".join(block for block in blocks if block)
 
 
+def render_execution_snapshot(
+    response: GetExecutionResponse,
+    *,
+    include_metadata: bool = True,
+    include_appendices: bool = True,
+    start_index: int = 0,
+    full_output: bool,
+) -> str:
+    blocks: list[str] = []
+    if include_metadata:
+        blocks.append(render_execution_header(response))
+
+    activity_blocks = render_activity_blocks(
+        response.activity, start_index=start_index, full_output=full_output
+    )
+    if activity_blocks:
+        blocks.append(activity_blocks)
+    elif include_metadata and response.activity_state == "unavailable":
+        blocks.append("No live activity is available yet.")
+
+    if include_appendices:
+        appendices = render_appendices(response)
+        if appendices:
+            blocks.append(appendices)
+
+    return "\n\n".join(blocks).rstrip()
+
+
 def render_appendices(response: GetExecutionResponse) -> str:
     appendices: list[str] = []
     if response.output_text:
@@ -112,6 +159,44 @@ def _render_activity(
     return f"{header}\n{body}"
 
 
+def _render_tool_group(
+    call: ToolCallActivity,
+    results: list[ToolOutputActivity | ToolErrorActivity],
+    *,
+    full_output: bool,
+    width: int,
+) -> str:
+    blocks = [_render_activity(call, full_output=full_output, width=width)]
+    for result in results:
+        blocks.append(
+            _indent_block(
+                _render_activity(result, full_output=full_output, width=width)
+            )
+        )
+    return "\n\n".join(blocks)
+
+
+def _group_tool_results(
+    activities: list[ExecutionActivityItem],
+) -> dict[str, list[ToolOutputActivity | ToolErrorActivity]]:
+    results: dict[str, list[ToolOutputActivity | ToolErrorActivity]] = {}
+    for activity in activities:
+        if not _is_tool_result(activity) or activity.tool_call_id is None:
+            continue
+        results.setdefault(activity.tool_call_id, []).append(activity)
+    return results
+
+
+def _is_tool_result(
+    activity: ExecutionActivityItem,
+) -> TypeGuard[ToolOutputActivity | ToolErrorActivity]:
+    return isinstance(activity, ToolOutputActivity | ToolErrorActivity)
+
+
+def _indent_block(block: str) -> str:
+    return "\n".join(f"  {line}" if line else line for line in block.splitlines())
+
+
 def _activity_header(
     activity: ExecutionActivityItem, *, tool_name: str | None, inline_meta: str | None
 ) -> str:

{alloy_runtime_cli-0.2.71 → alloy_runtime_cli-0.2.73}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "alloy-runtime-cli"
-version = "0.2.71"
+version = "0.2.73"
 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==0.2.71",
-  "alloy-runtime-types==0.2.71",
+  "alloy-runtime-sdk==0.2.73",
+  "alloy-runtime-types==0.2.73",
   "trogon==0.6.0",
 ]
 

alloy_runtime_cli-0.2.71/cli/commands/executions/get/presenter.py

@@ -1,49 +0,0 @@
-from alloy_runtime_types.dtos.executions import GetExecutionResponse
-from cli.commands.executions.get.transcript_renderer import (
-    render_activity_blocks,
-    render_appendices,
-    render_execution_header,
-)
-from cli.infrastructure.output import OutputFormat, OutputService
-
-
-def present_execution(
-    response: GetExecutionResponse,
-    *,
-    include_metadata: bool = True,
-    include_appendices: bool | None = None,
-    start_index: int = 0,
-    full_output: bool = False,
-) -> int:
-    output = OutputService.get()
-    if output.format == OutputFormat.JSON:
-        output.raw(response)
-        return len(response.activity)
-
-    blocks: list[str] = []
-    if include_metadata:
-        blocks.append(render_execution_header(response))
-
-    activity_blocks = render_activity_blocks(
-        response.activity,
-        start_index=start_index,
-        full_output=full_output,
-    )
-    if activity_blocks:
-        blocks.append(activity_blocks)
-    elif include_metadata and response.activity_state == "unavailable":
-        blocks.append("No live activity is available yet.")
-
-    should_render_appendices = (
-        include_metadata if include_appendices is None else include_appendices
-    )
-    if should_render_appendices:
-        appendices = render_appendices(response)
-        if appendices:
-            blocks.append(appendices)
-
-    rendered = "\n\n".join(blocks).rstrip()
-    if rendered:
-        output.text(rendered)
-
-    return len(response.activity)