refactor(plugin-eval): remove dead auth/model_tier config (#591)

plugins/plugin-eval/src/plugin_eval/cli.py

@@ -35,7 +35,6 @@ def _run_score(
     output: str,
     verbose: bool,
     concurrency: int,
-    auth: str,
     threshold: float | None,
 ) -> int:
     """Core scoring logic; returns exit code."""
@@ -48,7 +47,6 @@ def _run_score(
         output_format=output,
         verbose=verbose,
         concurrency=concurrency,
-        auth=auth,
     )
     engine = EvalEngine(config)
 
@@ -96,13 +94,12 @@ def score(
     output: str = typer.Option("markdown", help="Output format: json|markdown|html"),  # noqa: B008
     verbose: bool = typer.Option(False, "--verbose", "-v", help="Verbose output"),  # noqa: B008
     concurrency: int = typer.Option(4, help="Max concurrent LLM calls"),  # noqa: B008
-    auth: str = typer.Option("max", help="Auth mode: max|api-key"),  # noqa: B008
     threshold: float | None = typer.Option(  # noqa: B008
         None, help="Minimum score threshold; exit code 1 if below"
     ),
 ) -> None:
     """Evaluate a plugin or skill directory and report its quality score."""
-    exit_code = _run_score(path, depth, output, verbose, concurrency, auth, threshold)
+    exit_code = _run_score(path, depth, output, verbose, concurrency, threshold)
     if exit_code != 0:
         raise typer.Exit(code=exit_code)
 
@@ -113,11 +110,10 @@ def certify(
     output: str = typer.Option("markdown", help="Output format: json|markdown|html"),  # noqa: B008
     verbose: bool = typer.Option(False, "--verbose", "-v", help="Verbose output"),  # noqa: B008
     concurrency: int = typer.Option(4, help="Max concurrent LLM calls"),  # noqa: B008
-    auth: str = typer.Option("max", help="Auth mode: max|api-key"),  # noqa: B008
     threshold: float | None = typer.Option(None, help="Minimum score threshold"),  # noqa: B008
 ) -> None:
     """Certify a plugin or skill (runs at deep depth)."""
-    exit_code = _run_score(path, Depth.DEEP, output, verbose, concurrency, auth, threshold)
+    exit_code = _run_score(path, Depth.DEEP, output, verbose, concurrency, threshold)
     if exit_code != 0:
         raise typer.Exit(code=exit_code)
 

plugins/plugin-eval/src/plugin_eval/engine.py

@@ -83,7 +83,6 @@ class EvalEngine:
 
             judge_config = JudgeConfig(
                 judges=self.config.judges,
-                auth=self.config.auth,
                 concurrency=self.config.concurrency,
             )
             judge = JudgeAnalyzer(judge_config)
@@ -98,7 +97,6 @@ class EvalEngine:
                 mc_config = MonteCarloConfig(
                     n_runs=n_runs,
                     concurrency=self.config.concurrency,
-                    auth=self.config.auth,
                 )
                 mc = MonteCarloAnalyzer(mc_config)
 

plugins/plugin-eval/src/plugin_eval/layers/judge.py

@@ -147,9 +147,7 @@ def _measured_score(result: dict, key: str) -> float | None:
 @dataclass
 class JudgeConfig:
     judges: int = 1
-    auth: str = "max"
     concurrency: int = 4
-    model_tier: str = "auto"
 
 
 # ---------------------------------------------------------------------------

plugins/plugin-eval/src/plugin_eval/layers/monte_carlo.py

@@ -40,7 +40,6 @@ class MonteCarloConfig:
 
     n_runs: int = 50
     concurrency: int = 4
-    auth: str = "max"
     seed: int = 42
     progress_callback: Callable[[int, int], None] | None = None
 

plugins/plugin-eval/src/plugin_eval/models.py

@@ -36,11 +36,9 @@ class Depth(StrEnum):
 class EvalConfig(BaseModel):
     depth: Depth = Depth.STANDARD
     concurrency: int = Field(default=4, ge=1, le=20)
-    model_tier: str = "auto"
     output_format: str = "json"
     verbose: bool = False
     corpus_path: str | None = None
-    auth: str = "max"
     judges: int = Field(default=1, ge=1, le=5)
     monte_carlo_n: int | None = None
 

plugins/plugin-eval/tests/test_judge.py

@@ -59,7 +59,7 @@ class TestJudgeConfig:
     def test_default_config(self):
         config = JudgeConfig()
         assert config.judges == 1
-        assert config.auth == "max"
+        assert config.concurrency == 4
 
 
 class TestJudgeAnalyzer:

plugins/plugin-eval/tests/test_models.py

@@ -21,7 +21,6 @@ class TestEvalConfig:
         config = EvalConfig()
         assert config.depth == Depth.STANDARD
         assert config.concurrency == 4
-        assert config.auth == "max"
 
     def test_custom_config(self):
         config = EvalConfig(depth=Depth.DEEP, concurrency=8)