feat: error handling, retry with backoff, and circuit breaker

Add robust error handling to the evolution loop and LLM adapters:
- Retry utility with exponential backoff for transient errors (429, 5xx, timeouts)
- Per-call error isolation in evaluator and judge adapter
- Circuit breaker in EvolutionLoop (trips after N consecutive failures)
- CLI flags: --max-retries, --error-strategy (skip|retry|abort)
- Config fields: max_retries, retry_delay_base, circuit_breaker_threshold, error_strategy
- 16 new unit tests covering all error handling paths

Co-Authored-By: Paperclip <noreply@paperclip.ing>

src/prometheus/infrastructure/proposer_adapter.py

@@ -11,14 +11,22 @@ import dspy
 from prometheus.domain.entities import Prompt, Trajectory
 from prometheus.domain.ports import ProposerPort
 from prometheus.infrastructure.dspy_modules import InstructionProposer
+from prometheus.infrastructure.retry import retry_with_backoff
 
 
 class DSPyProposerAdapter(ProposerPort):
     """Uses evaluation trajectories to build a failure report and propose a new prompt."""
 
-    def __init__(self, lm: dspy.LM) -> None:
+    def __init__(
+        self,
+        lm: dspy.LM,
+        max_retries: int = 3,
+        retry_delay_base: float = 1.0,
+    ) -> None:
         self._lm = lm
         self._proposer = InstructionProposer()
+        self._max_retries = max_retries
+        self._retry_delay_base = retry_delay_base
 
     def propose(
         self,
@@ -27,13 +35,21 @@ class DSPyProposerAdapter(ProposerPort):
         task_description: str,
     ) -> Prompt:
         failure_examples = self._format_failures(trajectories)
-        with dspy.context(lm=self._lm):
-            pred = self._proposer(
-                current_instruction=current_prompt.text,
-                task_description=task_description,
-                failure_examples=failure_examples,
-            )
-        return Prompt(text=pred.new_instruction)
+
+        def _call() -> Prompt:
+            with dspy.context(lm=self._lm):
+                pred = self._proposer(
+                    current_instruction=current_prompt.text,
+                    task_description=task_description,
+                    failure_examples=failure_examples,
+                )
+            return Prompt(text=pred.new_instruction)
+
+        return retry_with_backoff(
+            _call,
+            max_retries=self._max_retries,
+            retry_delay_base=self._retry_delay_base,
+        )
 
     @staticmethod
     def _format_failures(trajectories: list[Trajectory]) -> str: