diff --git a/scripts/multi_agent.py b/scripts/multi_agent.py
index af5e6a1..adfc07e 100644
--- a/scripts/multi_agent.py
+++ b/scripts/multi_agent.py
@@ -50,6 +50,9 @@ def main():
                         help="Resume a SUSPENDED run: skip config_builder, poll "
                              "this experiment_id, then run the analyzer. The "
                              "goal is still required (the analyzer needs it).")
+    parser.add_argument("--output", metavar="FILE.md",
+                        help="Write the final analysis/answer to this markdown file "
+                             "(in addition to printing it).")
     parser.add_argument("--poll-budget-seconds", type=float, default=180.0,
                         help="How long the poll stage waits before suspending.")
     args = parser.parse_args()
@@ -75,6 +78,15 @@ def main():
     for stage, sub_id in result.sub_run_ids.items():
         console.print(f"[dim]  {stage}: {sub_id}[/dim]")
 
+    if args.output and result.analysis:
+        with open(args.output, "w", encoding="utf-8") as f:
+            f.write(result.analysis + "\n\n---\n"
+                    f"*Generated by {args.model} | orchestrator run "
+                    f"{result.orchestrator_run_id} | outcome: {result.outcome}. "
+                    "Agent-drafted from sealed capsules — verify numbers with "
+                    "scripts/tools/grade_analyses.py before publication.*\n")
+        console.print(f"[dim]Analysis written to {args.output}[/dim]")
+
     if result.outcome in ("complete", "answered", "needs_experiment") and result.analysis:
         console.print(Panel(Markdown(result.analysis), title="📊 Analysis",
                             border_style="green"))