From ba0da8c12217c2515a7a211f1cb4ab43fdf6488e Mon Sep 17 00:00:00 2001
From: "jinli.yl" <jinli.yl@alibaba-inc.com>
Date: Mon, 24 Aug 2026 17:35:14 +0800
Subject: [PATCH] fix(daily-paper): sanitize analysis workflow state

---
 reme/steps/cookbook/daily_paper/analyze.py | 9 +++++----
 tests/unit/test_daily_paper.py             | 8 +++++---
 2 files changed, 10 insertions(+), 7 deletions(-)

diff --git a/reme/steps/cookbook/daily_paper/analyze.py b/reme/steps/cookbook/daily_paper/analyze.py
index 54b9bd4a..68e3579f 100644
--- a/reme/steps/cookbook/daily_paper/analyze.py
+++ b/reme/steps/cookbook/daily_paper/analyze.py
@@ -128,8 +128,9 @@ class DailyPaperAnalyzeStep(DailyPaperStep):
         )
         used_titles.add(title)
         note_rel = note_path.relative_to(self.workspace_path).as_posix()
-        body = strip_frontmatter(output.body)
-        if not output.desc.strip() or not body:
+        desc = replace_surrogates(output.desc.strip())
+        body = replace_surrogates(strip_frontmatter(output.body))
+        if not desc or not body:
             raise ValueError(f"Agent returned an empty paper note for {paper.arxiv_id}")
         await write_markdown(
             note_path,
@@ -137,7 +138,7 @@ class DailyPaperAnalyzeStep(DailyPaperStep):
             {
                 "name": title,
                 "title": title,
-                "description": output.desc.strip(),
+                "description": desc,
                 "kind": "daily-paper-analysis",
                 "arxiv_id": paper.arxiv_id,
                 "source_title": paper.title,
@@ -165,7 +166,7 @@ class DailyPaperAnalyzeStep(DailyPaperStep):
             arxiv_id=paper.arxiv_id,
             reasoning=selected.reasoning,
             title=title,
-            desc=output.desc.strip(),
+            desc=desc,
             body=body,
             note_path=note_rel,
             pdf_path=pdf_rel,
diff --git a/tests/unit/test_daily_paper.py b/tests/unit/test_daily_paper.py
index 0583eec8..851a93da 100644
--- a/tests/unit/test_daily_paper.py
+++ b/tests/unit/test_daily_paper.py
@@ -778,8 +778,8 @@ async def test_pipeline_filters_strict_yesterday_and_writes_outputs(
             },
             {
                 "title": "记忆代理研究",
-                "desc": "Detailed note one",
-                "body": "Evidence one [p. 1].",
+                "desc": "Detailed note\ud800 one",
+                "body": "Evidence\udfff one [p. 1].",
             },
             {
                 "title": "上下文压缩研究",
@@ -874,7 +874,9 @@ async def test_pipeline_filters_strict_yesterday_and_writes_outputs(
     assert "ReMe" not in analysis_prompt
     assert "# PDF 分页文本" in analysis_prompt
     digest_prompt = cc_wrapper.calls[-1]["inputs"]
-    assert "Evidence one [p. 1]." in digest_prompt
+    assert "Evidence\ufffd one [p. 1]." in digest_prompt
+    assert "Detailed note\ufffd one" in digest_prompt
+    assert not any("\ud800" <= character <= "\udfff" for character in digest_prompt)
     assert "调用 Read" not in digest_prompt
     assert "daily/2026-07-21" not in digest_prompt
     assert "长期记忆" not in digest_prompt