diff --git a/offline-backend-20260512/backend/config.yaml b/offline-backend-20260512/backend/config.yaml
index ee7b9d07..1fd52073 100644
--- a/offline-backend-20260512/backend/config.yaml
+++ b/offline-backend-20260512/backend/config.yaml
@@ -425,6 +425,16 @@ ai_writing:
   keyword_model: deepseek-chat # 检索词生成专用模型，留空回退写作主模型
   rank_model: deepseek-chat # 素材排序专用模型，留空回退写作主模型
   max_materials: 20 # 合并去重后的素材总量上限，范围 [1, 100]，默认 20
+  # 用「真实智能体运行时」跑配置技能（read SKILL.md + 全套工具 + 沙箱），让技能的真实
+  # 检索逻辑（调内网接口/MCP/读文件）真正执行，而不是塞进通用 web/ES 检索 provider。
+  # 内网技能必须开 true，否则技能会走通用 provider、连不上内网而报 ConnectError。
+  # 仅对非「知识库检索(knowledge-base-search)」的配置技能生效；失败回退直连检索。
+  researcher_skill_agent: true
+  # 最终兜底：当所有技能都没检索到素材时，直连内网知识库 ES 接口兜底检索（避免素材全空）。
+  # 配上你内网 ES 接口地址即启用；留空则不启用该兜底。
+  # 接口契约：POST {url} 体 {"query","knowledge_base_name","ts":"yes"} → [{"page_content","metadata":{m_title,content,recUuid,m_publish,source1}}]
+  intranet_search_url: "" # 例如 http://10.106.57.180:7863/knowledge_base/search_docs_es
+  intranet_knowledge_base: "samples" # 兜底检索用的知识库名
   # 四阶段「内置功能型智能体」开关（按阶段灰度）。true = 该阶段节点改为加载
   # .deer-flow/agents/ai-writing-<stage>/ 的 SOUL.md + config（admin 可在智能体
   # 管理页编辑人设/模型），false / 缺失 = 走旧的硬编码 prompt 路径。
diff --git a/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/nodes/_skill_agent.py b/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/nodes/_skill_agent.py
index b89c9acd..b0951756 100644
--- a/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/nodes/_skill_agent.py
+++ b/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/nodes/_skill_agent.py
@@ -114,7 +114,25 @@ async def run_skill_via_real_agent(
         logger.warning("skill real-agent[%s]: build tools failed: %s", skill_name, e)
         return [], f"工具初始化失败: {e}"
 
-    system_prompt = (soul + "\n\n" if soul else "") + _MATERIAL_OUTPUT_CONTRACT
+    # 技能在容器内的目录：注入 SKILL.md 时模型只有内容、不知道脚本在哪，会用相对路径
+    # （python3 scripts/xxx.py）在错误的 CWD 下跑 → No such file。把绝对目录告诉它，让它
+    # 先 cd 进去再执行（与「别的智能体」里 skill_view/read_file 拿到 location 后的行为一致）。
+    try:
+        base_path = getattr(getattr(app_config, "skills", None), "container_path", None) or "/mnt/skills"
+    except Exception:  # noqa: BLE001
+        base_path = "/mnt/skills"
+    try:
+        skill_dir = skill.get_container_path(base_path)
+    except Exception:  # noqa: BLE001
+        skill_dir = ""
+    skill_loc_hint = (
+        f"\n\n【技能目录（重要）】技能「{skill_name}」在容器内的目录是：{skill_dir}\n"
+        f"- 运行该技能自带的脚本时，**务必先 `cd {skill_dir}` 再执行**"
+        f"（例如 `cd {skill_dir} && python3 scripts/xxx.py ...`），或对脚本用绝对路径；\n"
+        "- **不要用相对路径**（如 `python3 scripts/xxx.py`），否则当前目录不对会报 No such file。"
+    ) if skill_dir else ""
+
+    system_prompt = (soul + "\n\n" if soul else "") + _MATERIAL_OUTPUT_CONTRACT + skill_loc_hint
 
     task_lines = [f"写作意图：{intent or '(无)'}"]
     if user_query:
diff --git a/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/nodes/researcher.py b/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/nodes/researcher.py
index 59797efa..7f4afede 100644
--- a/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/nodes/researcher.py
+++ b/offline-backend-20260512/backend/packages/harness/deerflow/agents/ai_writing/nodes/researcher.py
@@ -725,7 +725,11 @@ async def researcher_node(state: AIWritingState, config: RunnableConfig) -> dict
                         user_query=effective_user_query if is_user_query_search else "",
                         config=config,
                         state=state,
-                        model_name_fallback=keyword_model,
+                        # 跑技能的 agent 用**用户选定的写作主模型**（state.model_name），不要用
+                        # keyword_model：那是「检索词生成快模型」，常配成外网模型（如 deepseek-chat
+                        # → api.deepseek.com），内网跑技能 agent 会连不上模型本身。传 None 让
+                        # resolve_agent_model_name 回退到 state.model_name（用户在表单选的、确定可达）。
+                        model_name_fallback=None,
                         max_results=per_skill,
                         timeout_s=_SKILL_SEARCH_TIMEOUT_SECONDS,
                         emit_query=_emit_q,
@@ -810,6 +814,43 @@ async def researcher_node(state: AIWritingState, config: RunnableConfig) -> dict
                         # 兜底已拿到素材即止，不再无谓尝试后续技能
                         break
 
+        # ── 最终兜底：技能（含兜底技能）都没检索到 → 直连内网知识库接口（ES）兜一把 ──
+        # 用户诉求：技能因网络/配置连不上而 0 条时，至少用内网 ES 接口直检，避免素材全空。
+        # 需在 config.yaml 的 ai_writing 段配 intranet_search_url 才启用；接口契约见
+        # IntranetSearchProvider（POST {query,knowledge_base_name,ts} → [{page_content,metadata}]）。
+        if not new_results:
+            intranet_url = (ai_writing_config.get("intranet_search_url") or "").strip()
+            if intranet_url:
+                from ..search.intranet_stub import IntranetSearchProvider
+
+                _emit_skill_step({
+                    "type": "fallback",
+                    "message": "技能均未检索到素材，改为直连内网知识库接口兜底检索…",
+                })
+                intranet = IntranetSearchProvider(
+                    base_url=intranet_url,
+                    knowledge_base_name=ai_writing_config.get("intranet_knowledge_base", "samples"),
+                )
+                fb_keywords = keywords[:keyword_count] or [intent]
+                fb_gathered = await asyncio.gather(
+                    *[intranet.search(kw, max_results=per_skill) for kw in fb_keywords],
+                    return_exceptions=True,
+                )
+                fb_added = 0
+                for fb_results in fb_gathered:
+                    if isinstance(fb_results, Exception):
+                        search_errors.append(f"{type(fb_results).__name__}: {fb_results}")
+                        continue
+                    fb_added += _collect(fb_results)
+                search_errors.extend(intranet.errors)
+                # 让下方 notice 拿到内网接口的原始响应体（解析失败时便于定位）
+                provider = intranet
+                _emit_skill_step({
+                    "type": "skill_result",
+                    "message": f"内网知识库接口兜底检索到 {fb_added} 条新素材",
+                    "matched_count": fb_added,
+                })
+
     # 合并去重
     # - 首次检索 / 普通 re_search（无输入，用户只想重出）：旧素材丢弃，覆盖式替换
     # - 素材不足求助 supplement_search：保留旧素材 + 追加新检索结果