fix(hypothesis): catch EmptyCompletionError dentro propose() invece di propagare

Run phase2-5-qwen25-prompt-mut-001 fallito perché qwen-2.5-72b ha emesso
empty completion ripetutamente; le 3 retry tenacity in LLMClient.complete
si sono esaurite e l'exception è bollata fino a run_phase1, marcando l'intero
run come failed (perso ~$0.017 di tier C).

Fix: HypothesisAgent.propose() ora cattura EmptyCompletionError nel loop
max_attempts trattandola come parse-fail "empty_completion" e ritentando
con max_retries+1 tentativi (cumulativo: max_retries tenacity × max_attempts
loop esterno, default 3×3 = 9 retry effettive). Se TUTTI i tentativi
falliscono con empty, ritorna proposal con strategy=None e
parse_error="empty_completion: ...", lasciando l'orchestrator continuare
con quel genome marcato come "fitness=0" senza crash totale del run.

+2 test: success dopo 1 empty + retry; failure totale dopo 3 empty consecutive.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This commit is contained in:
2026-05-12 00:07:53 +02:00
parent ec80af9f26
commit a12aead3e5
2 changed files with 44 additions and 3 deletions
+9 -2
View File
@@ -4,7 +4,7 @@ import re
from dataclasses import dataclass, field
from ..genome.hypothesis import HypothesisAgentGenome
from ..llm.client import CompletionResult, LLMClient
from ..llm.client import CompletionResult, EmptyCompletionError, LLMClient
from ..protocol.parser import ParseError, Strategy, parse_strategy
from ..protocol.validator import ValidationError, validate_strategy
@@ -274,7 +274,14 @@ class HypothesisAgent:
previous_error=errors[-1],
)
completion = self._llm.complete(genome, system=system, user=user)
try:
completion = self._llm.complete(genome, system=system, user=user)
except EmptyCompletionError as e:
# LLM esaurito retry tenacity senza una risposta. Tratta come
# parse-fail "empty" e ritenta nel loop esterno (max_attempts).
errors.append(f"empty_completion: {e}")
last_raw = ""
continue
completions.append(completion)
last_raw = completion.text
+35 -1
View File
@@ -2,7 +2,7 @@ import json
from multi_swarm.agents.hypothesis import HypothesisAgent, MarketSummary
from multi_swarm.genome.hypothesis import HypothesisAgentGenome, ModelTier
from multi_swarm.llm.client import CompletionResult
from multi_swarm.llm.client import CompletionResult, EmptyCompletionError
def make_summary() -> MarketSummary:
@@ -214,3 +214,37 @@ def test_hypothesis_agent_no_retry_when_first_succeeds(mocker): # type: ignore[
assert proposal.n_attempts == 1
assert len(proposal.completions) == 1
assert fake_llm.complete.call_count == 1
def test_hypothesis_agent_retries_on_empty_completion(mocker): # type: ignore[no-untyped-def]
"""LLMClient esaurisce retry tenacity → propose ritenta nel loop max_attempts."""
fake_llm = mocker.MagicMock()
fake_llm.complete.side_effect = [
EmptyCompletionError("empty response from qwen"),
CompletionResult(
text=VALID_STRATEGY_JSON,
input_tokens=200,
output_tokens=80,
tier=ModelTier.C,
model="qwen",
),
]
agent = HypothesisAgent(llm=fake_llm, max_retries=2)
proposal = agent.propose(make_genome(), make_summary())
assert proposal.strategy is not None
assert fake_llm.complete.call_count == 2
# n_attempts conta solo le completions arrivate (skipping empty failures).
assert len(proposal.completions) == 1
def test_hypothesis_agent_returns_failed_proposal_on_only_empty_completions(mocker): # type: ignore[no-untyped-def]
"""Tutti i tentativi sollevano EmptyCompletionError → proposal con strategy None."""
fake_llm = mocker.MagicMock()
fake_llm.complete.side_effect = EmptyCompletionError("empty response")
agent = HypothesisAgent(llm=fake_llm, max_retries=2)
proposal = agent.propose(make_genome(), make_summary())
assert proposal.strategy is None
assert proposal.parse_error is not None
assert "empty_completion" in proposal.parse_error
# 3 tentativi tutti falliti.
assert fake_llm.complete.call_count == 3