Add an LLM repair pass for malformed review outputs that miss the strict schema

[wwind123]

Summary\n\nWhen an agent produces a review or follow-up response that is substantively clear but fails the strict grammar/schema check, the orchestrator could invoke a smaller analyzer LLM to normalize the result into the expected structured shape.\n\nThe analyzer would read a small input: the raw agent output plus the exact schema contract and the current unresolved-item context. It would then return a strict structured object indicating:\n\n- approved or blocking\n- addressed items\n- remaining items\n- any new items introduced\n- optionally a short normalized summary or quoted evidence\n\n## Motivation\n\nRecent failures show the main issue is often not the underlying review content, but the model drifting away from the required wrapper format after a long turn. A focused repair pass may recover those near-miss outputs much more reliably than retrying the full review prompt.\n\n## Guardrails\n\n- The primary contract should remain structured output from the original agent.\n- The analyzer is only a fallback when strict parsing fails.\n- If the analyzer cannot confidently justify a repair, the orchestrator should treat the result as blocking rather than inventing approval.\n- The analyzer should not be able to override a clearly blocking review into approval without evidence from the raw output.\n\n## Relationship to #140\n\nThis is a follow-up idea to the structured-response migration in #140, not a replacement for it. Structured output should remain the primary path; the analyzer is only a recovery layer for malformed near-miss responses.\n\n## Example\n\nOn recent llm-dialectic runs, some reviews clearly contained the needed disposition information, but the model wrapped the JSON in fences or mixed in prose, causing the strict parser to fail. A repair pass could salvage those cases and reduce unnecessary reruns.\n\n-- OpenAI Codex