diff --git a/backend/agenteval/evaluation/campaign_runner.py b/backend/agenteval/evaluation/campaign_runner.py
index 46f02f2..3534fd2 100644
--- a/backend/agenteval/evaluation/campaign_runner.py
+++ b/backend/agenteval/evaluation/campaign_runner.py
@@ -20,6 +20,7 @@ from typing import Optional
 
 from sqlmodel import Session
 
+from agenteval.evaluation.analysis import resolve_analysis_model, start_campaign_analysis
 from agenteval.evaluation.campaign_scheduler import (
     TickAction,
     clock_offset,
@@ -183,6 +184,23 @@ async def advance_campaign(
     return result
 
 
+def _auto_start_analysis(campaign: Campaign, session: Session) -> None:
+    """Enqueue 智能分析 when a 正式线 campaign completes.
+
+    Accelerated campaigns, an unresolvable analysis model, and any enqueue
+    failure all skip silently — the analysis is an enhancement and must never
+    block or break campaign completion.
+    """
+    if campaign.time_scale != 1 or not campaign.id:
+        return
+    try:
+        if resolve_analysis_model(campaign, session) is None:
+            return
+        start_campaign_analysis(campaign.id, triggered_by="auto")
+    except Exception as exc:
+        _logger.warning("活动 %s 自动分析触发失败（已跳过）: %s", campaign.id, exc)
+
+
 # ── durable scheduler loop ──────────────────────────────────────────────────
 
 
@@ -232,6 +250,7 @@ async def run_campaign_loop(
                     current.status = CampaignStatus.COMPLETED
                     current.completed_at = utc_now()
                     repo.update(current)
+                    _auto_start_analysis(current, session)
                 return
 
             try:
diff --git a/backend/agenteval/evaluation/report_render.py b/backend/agenteval/evaluation/report_render.py
index 1dbd6f4..12ebac9 100644
--- a/backend/agenteval/evaluation/report_render.py
+++ b/backend/agenteval/evaluation/report_render.py
@@ -173,8 +173,57 @@ def render_markdown(report: dict[str, Any]) -> str:
     return "\n".join(lines)
 
 
-def render_campaign_markdown(report: dict[str, Any]) -> str:
-    """Render a dual-axis campaign report dict as Markdown."""
+_SEVERITY_LABELS = {"high": "高", "medium": "中", "low": "低"}
+
+
+def _render_analysis_lines(analysis: dict[str, Any], scenario_names: dict[str, str]) -> list[str]:
+    """Appendix for the campaign export: the structured analysis blocks."""
+
+    def _name(sid: str) -> str:
+        return scenario_names.get(sid, sid[:8])
+
+    lines = ["", "## 智能分析", "", "### 总体结论", "", str(analysis.get("overall") or "—"), ""]
+    problems = analysis.get("problems") or []
+    if problems:
+        lines += ["### 问题诊断", ""]
+        for p in problems:
+            severity = _SEVERITY_LABELS.get(p.get("severity"), "中")
+            names = "、".join(_name(sid) for sid in p.get("scenario_ids") or [])
+            head = f"- **[{severity}] {p.get('title', '')}**"
+            if names:
+                head += f"（场景：{names}）"
+            lines.append(head)
+            if p.get("description"):
+                lines.append(f"  {p['description']}")
+            evidence = p.get("evidence_run_ids") or []
+            if evidence:
+                lines.append("  证据 run：" + " ".join(f"`{rid}`" for rid in evidence))
+        lines.append("")
+    narratives = analysis.get("scenario_narratives") or []
+    if narratives:
+        lines += ["### 分场景叙述", ""]
+        for n in narratives:
+            lines.append(f"- **{_name(str(n.get('scenario_id', '')))}**：{n.get('narrative', '')}")
+        lines.append("")
+    suggestions = sorted(analysis.get("suggestions") or [], key=lambda s: s.get("priority", 0))
+    if suggestions:
+        lines += ["### 改善建议", ""]
+        lines += [f"{i}. {s.get('text', '')}" for i, s in enumerate(suggestions, 1)]
+        lines.append("")
+    return lines
+
+
+def render_campaign_markdown(
+    report: dict[str, Any],
+    *,
+    analysis: Optional[dict[str, Any]] = None,
+    scenario_names: Optional[dict[str, str]] = None,
+) -> str:
+    """Render a dual-axis campaign report dict as Markdown.
+
+    ``analysis`` is the stored 智能分析 result dict (completed only); when
+    absent the export is identical to the pre-analysis format.
+    """
     s = report["summary"]
     lines: list[str] = [
         f"# 活动周期报告 — {report['name']}",
@@ -216,4 +265,6 @@ def render_campaign_markdown(report: dict[str, Any]) -> str:
             f"| {c['scenario_name']} | {c['run_count']} | {_pct(c['pass_rate'])} | "
             f"{_pct(c['availability'])} | {_ms(c['avg_latency_ms'])} |"
         )
+    if analysis:
+        lines += _render_analysis_lines(analysis, scenario_names or {})
     return "\n".join(lines)
diff --git a/backend/agenteval/web/routers/campaigns.py b/backend/agenteval/web/routers/campaigns.py
index 88ecfe7..ebde8dc 100644
--- a/backend/agenteval/web/routers/campaigns.py
+++ b/backend/agenteval/web/routers/campaigns.py
@@ -124,7 +124,13 @@ async def get_campaign_report_markdown(campaign_id: str, session: Session = Depe
         raise HTTPException(status_code=404, detail="campaign not found")
     runs = RunRepository(session).list_by_campaign(campaign_id)
     scenario_names = {s.id: s.name for s in ScenarioRepository(session).list_all()}
-    md = render_campaign_markdown(generate_campaign_report(campaign, runs, scenario_names=scenario_names))
+    analysis_row = CampaignAnalysisRepository(session).get_by_campaign(campaign_id)
+    analysis = analysis_row.get_result() if analysis_row and analysis_row.status == "completed" else None
+    md = render_campaign_markdown(
+        generate_campaign_report(campaign, runs, scenario_names=scenario_names),
+        analysis=analysis,
+        scenario_names=scenario_names,
+    )
     return Response(
         content=md,
         media_type="text/markdown; charset=utf-8",
diff --git a/tests/integration/test_campaign_analysis_api.py b/tests/integration/test_campaign_analysis_api.py
index 43aff15..f54ea59 100644
--- a/tests/integration/test_campaign_analysis_api.py
+++ b/tests/integration/test_campaign_analysis_api.py
@@ -183,3 +183,30 @@ async def test_rerun_upserts_without_new_row(client, seeded_db, monkeypatch):
 async def test_get_analysis_missing_campaign_404(client, seeded_db):
     resp = await client.get("/api/campaigns/nope/analysis")
     assert resp.status_code == 404
+
+
+async def test_markdown_export_includes_completed_analysis(client, seeded_db):
+    campaign_id = await _create_campaign(client, seeded_db)
+    _complete_analysis_row(seeded_db, campaign_id)
+    resp = await client.get(f"/api/campaigns/{campaign_id}/report/markdown")
+    assert resp.status_code == 200
+    assert "## 智能分析" in resp.text
+    assert "整体达标" in resp.text
+    assert "表现稳定" in resp.text
+
+
+async def test_markdown_export_ignores_non_completed_analysis(client, seeded_db):
+    campaign_id = await _create_campaign(client, seeded_db)
+    row = CampaignAnalysisDB(campaign_id=campaign_id, status="failed", error="模型超时")
+    seeded_db.add(row)
+    seeded_db.commit()
+    resp = await client.get(f"/api/campaigns/{campaign_id}/report/markdown")
+    assert resp.status_code == 200
+    assert "智能分析" not in resp.text
+
+
+async def test_markdown_export_without_analysis_unchanged(client, seeded_db):
+    campaign_id = await _create_campaign(client, seeded_db)
+    resp = await client.get(f"/api/campaigns/{campaign_id}/report/markdown")
+    assert resp.status_code == 200
+    assert "智能分析" not in resp.text
diff --git a/tests/integration/test_campaign_analysis_auto_trigger.py b/tests/integration/test_campaign_analysis_auto_trigger.py
new file mode 100644
index 0000000..9fa5add
--- /dev/null
+++ b/tests/integration/test_campaign_analysis_auto_trigger.py
@@ -0,0 +1,138 @@
+"""Integration tests for the runner's auto analysis hook (v0.7 ticket 05).
+
+Only 正式线 campaigns (time_scale == 1) enqueue the analysis task on
+COMPLETED; accelerated/cancelled campaigns and missing analysis models all
+skip silently. The analysis service itself is spied, not executed.
+"""
+
+import pytest
+from agenteval.evaluation import campaign_runner
+from agenteval.evaluation.campaign_runner import request_cancel, start_campaign
+from agenteval.models import (
+    Campaign,
+    CampaignPlanEntry,
+    CampaignStatus,
+    Case,
+    CaseType,
+    ChannelType,
+    EvalTarget,
+    PlatformType,
+    Scenario,
+    TargetStatus,
+)
+from agenteval.storage.db import utc_now
+from agenteval.storage.repository import CampaignRepository, ScenarioRepository, TargetRepository
+
+from tests.unit.mock_channel import MockChannel
+
+TICK = 0.01
+
+
+@pytest.fixture()
+def seeded_db(db_session, monkeypatch):
+    from agenteval.channels import factory as factory_module
+    from agenteval.evaluation import engine as engine_module
+    from agenteval.storage import db as db_module
+    from agenteval.storage import repository as repo_module
+
+    def _test_get_session():
+        return db_session
+
+    monkeypatch.setattr(db_module, "get_session", _test_get_session)
+    monkeypatch.setattr(repo_module, "get_session", _test_get_session)
+    monkeypatch.setattr(engine_module, "get_session", _test_get_session)
+    monkeypatch.setattr(campaign_runner, "get_session", _test_get_session)
+
+    channel = MockChannel(reply_delay=0.0)
+    monkeypatch.setattr(factory_module.ChannelFactory, "create", lambda target: channel)
+
+    TargetRepository(db_session).create(EvalTarget(
+        id="t-1", name="mock-target",
+        platform=PlatformType.AI_DIGITAL_EMPLOYEE, channel_type=ChannelType.TUTU_API,
+        channel_config={"base_url": "http://mock", "token": "x"}, status=TargetStatus.ACTIVE,
+    ))
+    ScenarioRepository(db_session).create(Scenario(
+        id="s-1", name="mock-scenario",
+        cases=[Case(id="c1", type=CaseType.SINGLE, messages=["hi"])],
+    ))
+    return db_session
+
+
+@pytest.fixture()
+def analysis_spy(monkeypatch):
+    """Spy the analysis seam: resolvable model, recorded enqueue calls."""
+    calls: list[tuple[str, str]] = []
+    monkeypatch.setattr(
+        campaign_runner, "start_campaign_analysis",
+        lambda cid, *, triggered_by: calls.append((cid, triggered_by)),
+    )
+    monkeypatch.setattr(campaign_runner, "resolve_analysis_model", lambda campaign, session: object())
+    return calls
+
+
+def _make_campaign(session, **overrides) -> Campaign:
+    payload = dict(
+        name="auto-trigger",
+        target_id="t-1",
+        window_seconds=1,
+        time_scale=1.0,  # 正式线：1s 窗口真实耗时 ~1s
+        plan=[CampaignPlanEntry(scenario_id="s-1", offset_seconds=0, count=1)],
+    )
+    payload.update(overrides)
+    return CampaignRepository(session).create(Campaign(**payload))
+
+
+async def _await_task(campaign_id, timeout=5.0):
+    import asyncio
+    task = campaign_runner.campaign_registry.get(campaign_id)
+    if task is not None:
+        await asyncio.wait_for(task, timeout=timeout)
+
+
+async def test_realtime_completion_auto_enqueues_analysis(seeded_db, analysis_spy):
+    campaign = _make_campaign(seeded_db)
+    start_campaign(campaign.id, seeded_db, tick_seconds=TICK)
+    await _await_task(campaign.id)
+
+    final = CampaignRepository(seeded_db).get(campaign.id)
+    assert final.status == CampaignStatus.COMPLETED
+    assert analysis_spy == [(campaign.id, "auto")]
+
+
+async def test_accelerated_completion_does_not_enqueue(seeded_db, analysis_spy):
+    campaign = _make_campaign(seeded_db, time_scale=1000.0)  # 加速调试线
+    start_campaign(campaign.id, seeded_db, tick_seconds=TICK)
+    await _await_task(campaign.id)
+
+    final = CampaignRepository(seeded_db).get(campaign.id)
+    assert final.status == CampaignStatus.COMPLETED
+    assert analysis_spy == []
+
+
+async def test_cancelled_campaign_does_not_enqueue(seeded_db, analysis_spy):
+    import asyncio
+    campaign = _make_campaign(seeded_db, window_seconds=100)
+    start_campaign(campaign.id, seeded_db, tick_seconds=TICK)
+    await asyncio.sleep(0.05)
+
+    repo = CampaignRepository(seeded_db)
+    current = repo.get(campaign.id)
+    current.status = CampaignStatus.CANCELLED
+    current.completed_at = utc_now()
+    repo.update(current)
+    request_cancel(campaign.id)
+    await _await_task(campaign.id)
+
+    assert repo.get(campaign.id).status == CampaignStatus.CANCELLED
+    assert analysis_spy == []
+
+
+async def test_missing_analysis_model_skips_silently(seeded_db, monkeypatch, analysis_spy):
+    monkeypatch.setattr(campaign_runner, "resolve_analysis_model", lambda campaign, session: None)
+    campaign = _make_campaign(seeded_db)
+    start_campaign(campaign.id, seeded_db, tick_seconds=TICK)
+    await _await_task(campaign.id)
+
+    final = CampaignRepository(seeded_db).get(campaign.id)
+    assert final.status == CampaignStatus.COMPLETED  # 活动完成流程不受影响
+    assert analysis_spy == []
diff --git a/tests/unit/test_report_render.py b/tests/unit/test_report_render.py
index 817f0f1..19f245c 100644
--- a/tests/unit/test_report_render.py
+++ b/tests/unit/test_report_render.py
@@ -185,3 +185,64 @@ def test_render_campaign_markdown_empty_bucket_dashes():
     md = render_campaign_markdown(_campaign_report())
     # bucket 1 has no runs: pass_rate/availability/latency all render as —
     assert "| 3600–7200 | 0 | — | — | — |" in md
+
+
+def _analysis() -> dict:
+    """A hand-built dict matching the stored campaign analysis result shape."""
+    return {
+        "overall": "整窗通过率偏低，售后场景拖后腿",
+        "problems": [
+            {
+                "severity": "high",
+                "title": "售后答非所问",
+                "description": "多轮对话中反复偏离用户问题",
+                "scenario_ids": ["s-1"],
+                "evidence_run_ids": ["run-abc", "run-def"],
+            },
+            {
+                "severity": "low",
+                "title": "响应偏慢",
+                "description": "高峰时段时延偏高",
+                "scenario_ids": ["s-2"],
+                "evidence_run_ids": [],
+            },
+        ],
+        "scenario_narratives": [
+            {"scenario_id": "s-1", "narrative": "售后场景表现不稳定"},
+            {"scenario_id": "s-2", "narrative": "售前场景表现稳定"},
+        ],
+        "suggestions": [
+            {"priority": 2, "text": "次要建议：扩容"},
+            {"priority": 1, "text": "首要建议：补充售后知识库"},
+        ],
+    }
+
+
+def test_render_campaign_markdown_appends_analysis_sections():
+    md = render_campaign_markdown(
+        _campaign_report(),
+        analysis=_analysis(),
+        scenario_names={"s-1": "售后场景", "s-2": "售前场景"},
+    )
+    assert "## 智能分析" in md
+    assert "### 总体结论" in md
+    assert "整窗通过率偏低，售后场景拖后腿" in md
+    assert "### 问题诊断" in md
+    assert "**[高] 售后答非所问**（场景：售后场景）" in md
+    assert "`run-abc`" in md and "`run-def`" in md
+    assert "**[低] 响应偏慢**（场景：售前场景）" in md
+    assert "### 分场景叙述" in md
+    assert "**售后场景**：售后场景表现不稳定" in md
+    assert "### 改善建议" in md
+    # 建议按 priority 升序
+    assert md.index("首要建议") < md.index("次要建议")
+
+
+def test_render_campaign_markdown_analysis_falls_back_to_id_prefix():
+    md = render_campaign_markdown(_campaign_report(), analysis=_analysis(), scenario_names={})
+    assert "**[高] 售后答非所问**（场景：s-1）" in md
+
+
+def test_render_campaign_markdown_without_analysis_unchanged():
+    md = render_campaign_markdown(_campaign_report())
+    assert "智能分析" not in md