Add consensus failure handling with fallback options for multi-agent pipelines

Implements detection and recovery for when agents fail to reach consensus:
- Orchestrator exits with code 2 on consensus failure (distinct from error=1)
- Records failed run context (proposals, agent states, conflicts) to Dragonfly
- Provides fallback options: rerun same, rerun with GAMMA, escalate tier, accept partial
- Adds UI alert with action buttons for user-driven recovery
- Adds failure details modal and downloadable failure report
- Only marks pipeline complete when consensus achieved or user accepts fallback

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

agents/multi-agent/orchestrator.ts

@@ -448,6 +448,27 @@ The solution should consider fault tolerance, data consistency, and cost optimiz
 
     analyzePerformance(metrics);
 
+    // Output special marker for server to parse consensus status
+    // Format: ORCHESTRATION_RESULT:{"consensus":true/false,"metrics":{...}}
+    console.log("\nORCHESTRATION_RESULT:" + JSON.stringify({
+      consensus: metrics.final_consensus,
+      task_id: metrics.task_id,
+      metrics: metrics
+    }));
+
+    // Exit with code 2 for consensus failure (distinct from error=1, success=0)
+    if (!metrics.final_consensus) {
+      console.log("\n[ORCHESTRATOR] Consensus NOT achieved - exiting with code 2");
+      exitCode = 2;
+
+      // Report consensus failure to observability
+      const pipelineId = process.env.PIPELINE_ID;
+      if (pipelineId) {
+        await reportErrorToObservability(pipelineId, "consensus_failure", "high",
+          `Agents failed to reach consensus. Conflicts: ${metrics.conflicts_detected}, Resolved: ${metrics.conflicts_resolved}`);
+      }
+    }
+
   } catch (e: any) {
     console.error("Orchestrator error:", e.message);
     exitCode = 1;