fluxloop-cli 0.2.23__tar.gz → 0.2.25__tar.gz

{fluxloop_cli-0.2.23 → fluxloop_cli-0.2.25}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: fluxloop-cli
-Version: 0.2.23
+Version: 0.2.25
 Summary: FluxLoop CLI for running agent simulations
 Author-email: FluxLoop Team <team@fluxloop.dev>
 License: Apache-2.0
@@ -26,6 +26,7 @@ Requires-Dist: httpx>=0.24.0
 Requires-Dist: rich>=13.0
 Requires-Dist: python-dotenv>=1.0.0
 Requires-Dist: fluxloop>=0.1.0
+Requires-Dist: ruamel.yaml>=0.17.0
 Provides-Extra: dev
 Requires-Dist: pytest>=7.0; extra == "dev"
 Requires-Dist: pytest-asyncio>=0.21.0; extra == "dev"

{fluxloop_cli-0.2.23 → fluxloop_cli-0.2.25}/fluxloop_cli/__init__.py

@@ -2,7 +2,7 @@
 FluxLoop CLI - Command-line interface for running agent simulations.
 """
 
-__version__ = "0.2.23"
+__version__ = "0.2.25"
 
 from .main import app
 

{fluxloop_cli-0.2.23 → fluxloop_cli-0.2.25}/fluxloop_cli/commands/record.py

@@ -8,6 +8,7 @@ from typing import Dict, Optional
 import typer
 from rich.console import Console
 from rich.table import Table
+from ruamel.yaml import YAML, CommentedMap
 
 from ..constants import DEFAULT_ROOT_DIR_NAME
 from ..project_paths import (
@@ -20,6 +21,9 @@ from ..config_schema import CONFIG_SECTION_FILENAMES
 
 app = typer.Typer()
 console = Console()
+_yaml = YAML()
+_yaml.indent(mapping=2, sequence=4, offset=2)
+_yaml.preserve_quotes = True
 
 
 def _load_env(env_path: Path) -> Dict[str, str]:
@@ -44,22 +48,25 @@ def _update_simulation(recording_enabled: bool, project: Optional[str], root: Pa
     except KeyError:
         return
 
-    data: Dict[str, object] = {}
     if simulation_path.exists():
-        import yaml
+        with simulation_path.open("r", encoding="utf-8") as handle:
+            loaded = _yaml.load(handle) or CommentedMap()
+    else:
+        loaded = CommentedMap()
 
-        data = yaml.safe_load(simulation_path.read_text()) or {}
+    if not isinstance(loaded, CommentedMap):
+        loaded = CommentedMap(loaded)
 
-    replay = data.setdefault("replay_args", {})
-    replay["enabled"] = recording_enabled
+    replay = loaded.get("replay_args")
+    if not isinstance(replay, CommentedMap):
+        replay = CommentedMap(replay or {})
+        loaded["replay_args"] = replay
 
-    import yaml
+    replay["enabled"] = recording_enabled
 
     simulation_path.parent.mkdir(parents=True, exist_ok=True)
-    simulation_path.write_text(
-        yaml.dump(data, sort_keys=False, default_flow_style=False),
-        encoding="utf-8",
-    )
+    with simulation_path.open("w", encoding="utf-8") as handle:
+        _yaml.dump(loaded, handle)
 
 
 @app.command()

{fluxloop_cli-0.2.23 → fluxloop_cli-0.2.25}/fluxloop_cli/conversation_supervisor.py

@@ -9,6 +9,7 @@ next user utterance consistent with the experiment persona and service context.
 from __future__ import annotations
 
 import json
+import logging
 import os
 import re
 from dataclasses import dataclass
@@ -18,6 +19,7 @@ import httpx
 
 from fluxloop.schemas import MultiTurnSupervisorConfig
 
+logger = logging.getLogger(__name__)
 SupervisorDecisionType = Literal["continue", "terminate"]
 
 
@@ -118,6 +120,13 @@ class ConversationSupervisor:
         )
 
         provider = (self.config.provider or "openai").lower()
+        logger.debug(
+            "supervisor.decide provider=%s turns=%d persona=%r service=%r",
+            provider,
+            len(turns),
+            persona_description,
+            service_context,
+        )
         if provider == "mock":
             return self._mock_decision(conversation_state)
         if provider == "openai":
@@ -131,6 +140,11 @@ class ConversationSupervisor:
         """Deterministic supervisor used for tests, scripted runs, or offline modes."""
 
         metadata = self.config.metadata or {}
+        logger.debug(
+            "mock supervisor state: scripted_questions=%s context_keys=%s",
+            isinstance(metadata.get("scripted_questions"), list),
+            list(conversation_state.get("context", {}).keys()),
+        )
 
         scripted_questions = metadata.get("scripted_questions")
         if isinstance(scripted_questions, list):

{fluxloop_cli-0.2.23 → fluxloop_cli-0.2.25}/fluxloop_cli/evaluation/engine/core.py

@@ -13,6 +13,13 @@ from typing import Any, Dict, List, Optional, Tuple, Literal
 
 import shutil
 from rich.console import Console
+from rich.progress import (
+    BarColumn,
+    Progress,
+    SpinnerColumn,
+    TextColumn,
+    TimeRemainingColumn,
+)
 
 from ..config import AggregateConfig, EvaluationConfig, EvaluatorConfig
 from ..llm import LLMEvaluationManager, LLMResult
@@ -321,9 +328,26 @@ def run_evaluation(
             output_dir=options.output_dir,
         )
 
-    for trace in traces:
-        outcome = _evaluate_trace(trace, config, llm_manager)
-        results.append(outcome)
+    console.print("\n[bold green]🧪 Evaluating traces...[/bold green]\n")
+
+    with Progress(
+        SpinnerColumn(),
+        TextColumn("[progress.description]{task.description}"),
+        BarColumn(),
+        TextColumn("[progress.percentage]{task.percentage:>3.0f}%"),
+        TimeRemainingColumn(),
+        TextColumn("({task.completed} of {task.total})"),
+        console=console,
+    ) as progress:
+        task_id = progress.add_task(
+            "Evaluating traces",
+            total=len(traces),
+        )
+
+        for trace in traces:
+            outcome = _evaluate_trace(trace, config, llm_manager)
+            results.append(outcome)
+            progress.advance(task_id)
 
     _write_per_trace(results, options.output_dir)
 

{fluxloop_cli-0.2.23 → fluxloop_cli-0.2.25}/fluxloop_cli/runner.py

@@ -7,6 +7,7 @@ from __future__ import annotations
 import asyncio
 import contextvars
 import json
+import logging
 import inspect
 import os
 import time
@@ -27,6 +28,7 @@ from .arg_binder import ArgBinder
 from .conversation_supervisor import ConversationSupervisor, SupervisorDecision
 
 console = Console()
+logger = logging.getLogger(__name__)
 
 
 class ExperimentRunner:
@@ -499,6 +501,13 @@ class ExperimentRunner:
                 while True:
                     callback_messages: Dict[str, Any] = {}
 
+                    logger.debug(
+                        "multi-turn request: iteration=%s turn=%s input_type=%s preview=%r",
+                        iteration,
+                        turn_count,
+                        type(current_user_input).__name__,
+                        current_user_input if isinstance(current_user_input, str) else str(current_user_input),
+                    )
                     result = await self._call_agent(
                         agent_func,
                         current_user_input,
@@ -524,6 +533,13 @@ class ExperimentRunner:
                         coerced = await self._coerce_result_to_text(result)
                         assistant_output = coerced if coerced is not None else str(result)
 
+                    logger.debug(
+                        "multi-turn agent output: raw_type=%s assistant_output_type=%s preview=%r",
+                        type(result).__name__,
+                        type(assistant_output).__name__,
+                        assistant_output if isinstance(assistant_output, str) else str(assistant_output),
+                    )
+
                     final_output = assistant_output
 
                     conversation_state["turns"].append(
@@ -547,6 +563,14 @@ class ExperimentRunner:
                         service_context=service_context,
                     )
                     last_decision = decision
+                    logger.debug(
+                        "multi-turn supervisor decision: decision=%s termination=%r next_type=%s",
+                        decision.decision,
+                        decision.termination_reason,
+                        type(decision.next_user_message).__name__
+                        if decision.next_user_message is not None
+                        else None,
+                    )
 
                     if decision.decision == "terminate":
                         termination_reason = (
@@ -569,6 +593,11 @@ class ExperimentRunner:
                             "Supervisor decided to continue but provided no next_user_message."
                         )
 
+                    logger.debug(
+                        "multi-turn next user message: type=%s preview=%r",
+                        type(next_user_message).__name__,
+                        next_user_message if isinstance(next_user_message, str) else str(next_user_message),
+                    )
                     conversation_state["turns"].append(
                         {
                             "role": "user",

{fluxloop_cli-0.2.23 → fluxloop_cli-0.2.25}/fluxloop_cli/templates.py

@@ -155,7 +155,7 @@ def create_simulation_config(project_name: str) -> str:
               When continuing, craft the next user message consistent with the persona.
               When terminating, explain the reason and provide any closing notes.
             metadata:
-              scripted_questions: []  # 순차 재생용 사용자 발화 배열 (예: ["첫 질문", "둘째 질문", ...])
+              scripted_questions: []  # Array of user utterances for sequential playback (e.g., ["First question", "Second question", ...])
               mock_decision: terminate        # Fallback when no scripted questions remain
               mock_reason: script_complete    # Termination reason for scripted runs
               mock_closing: "Thanks for the help. I have no further questions."

{fluxloop_cli-0.2.23 → fluxloop_cli-0.2.25}/fluxloop_cli.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: fluxloop-cli
-Version: 0.2.23
+Version: 0.2.25
 Summary: FluxLoop CLI for running agent simulations
 Author-email: FluxLoop Team <team@fluxloop.dev>
 License: Apache-2.0
@@ -26,6 +26,7 @@ Requires-Dist: httpx>=0.24.0
 Requires-Dist: rich>=13.0
 Requires-Dist: python-dotenv>=1.0.0
 Requires-Dist: fluxloop>=0.1.0
+Requires-Dist: ruamel.yaml>=0.17.0
 Provides-Extra: dev
 Requires-Dist: pytest>=7.0; extra == "dev"
 Requires-Dist: pytest-asyncio>=0.21.0; extra == "dev"

{fluxloop_cli-0.2.23 → fluxloop_cli-0.2.25}/fluxloop_cli.egg-info/requires.txt

@@ -5,6 +5,7 @@ httpx>=0.24.0
 rich>=13.0
 python-dotenv>=1.0.0
 fluxloop>=0.1.0
+ruamel.yaml>=0.17.0
 
 [anthropic]
 anthropic>=0.7.0

{fluxloop_cli-0.2.23 → fluxloop_cli-0.2.25}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "fluxloop-cli"
-version = "0.2.23"
+version = "0.2.25"
 description = "FluxLoop CLI for running agent simulations"
 authors = [
     {name = "FluxLoop Team", email = "team@fluxloop.dev"}
@@ -32,6 +32,7 @@ dependencies = [
     "rich>=13.0",
     "python-dotenv>=1.0.0",
     "fluxloop>=0.1.0",
+    "ruamel.yaml>=0.17.0",
 ]
 
 [project.optional-dependencies]