fix(artifacts): keep sandbox paths inside CJK corner brackets whole (#7377)

The quote-delimiter tables only knew ASCII quotes, backticks and curly
quotes, so `已生成「/mnt/user-data/outputs/季度 报告.pdf」。` captured the
handle `/mnt/user-data/outputs/季度` (the path pattern stops at the space)
and `已生成『…』。` kept the closer inside `real_ref`; either handle is dead
on a later read.

Register `「」` and `『』` as wrappers: a paired closer stays inside a
filename, an unpaired one is trimmed from a bare token, and corner-trailed
URLs and refs are no longer cut.

The persistence regression (capture -> registry round-trip -> resolution ->
real file read with the tool result gone) is ported from #7376 by
@yangjj-iso, per review.

backend/packages/harness/deerflow/tools/AGENTS.md

@@ -6,10 +6,10 @@ shape, even when invalid; never mix its MIME metadata with a conflicting legacy
 source. Both shapes use the existing referenceability gate and per-result deduplication.
 
 Free-text sandbox paths retain spaces inside matching ASCII/curly quote pairs,
-backticks or backslash-escaped wrappers. Exclude wrapper escapes without stripping
-literal path backslashes. Reject Python line boundaries and tabs; bound delimiter
-searches to the same line so later refs survive. Preserve bare-token boundaries
-and structured refs. Tests: `tests/test_artifact_registry.py`.
+CJK corner brackets, backticks or backslash-escaped wrappers. Exclude wrapper
+escapes without stripping literal path backslashes. Reject Python line boundaries
+and tabs; bound delimiter searches to the same line so later refs survive.
+Preserve bare-token boundaries and structured refs. Tests: `tests/test_artifact_registry.py`.
 
 `artifact_registry.py` checks remote file suffixes against the complete parsed
 URL path without case sensitivity. Free-text URL and sandbox-path extraction

backend/packages/harness/deerflow/tools/artifact_registry.py

@@ -34,7 +34,7 @@ _SANDBOX_PATH_PATTERN = re.compile(r"/mnt/user-data/\S+")
 # Delimiters a tool result may wrap a path in. A delimited span is one value, so a
 # filename inside it keeps its spaces; the bare token scan below keeps its
 # whitespace boundary, because nothing tells it where such a name would end.
-_REF_QUOTE_DELIMITERS = {'"': '"', "'": "'", "`": "`", "“": "”", "‘": "’"}
+_REF_QUOTE_DELIMITERS = {'"': '"', "'": "'", "`": "`", "“": "”", "‘": "’", "「": "」", "『": "』"}
 _REF_LINE_BOUNDARY_PATTERN = re.compile(r"[\r\n\v\f\x1c-\x1e\x85\u2028\u2029]")
 _SPAN_BARE = "bare"
 _SPAN_QUOTED = "quoted"
@@ -57,8 +57,8 @@ _STRUCTURED_TASK_KEYS = frozenset({"task_id", "job_id"})
 # quotes/brackets/backticks that markdown- and JSON-formatted tool output
 # commonly wraps paths in, including CJK punctuation. `\S+` would otherwise
 # consume them into `real_ref`.
-_REF_TRAILING_NOISE_CHARS = ".,;:)]}\"'`\u3002\uff0c\uff1b\uff1a\u3001\uff09\u3011\u300b\u201d\u2019"
-_REF_CLOSING_DELIMITERS = {")": "(", "]": "[", "}": "{", "）": "（", "】": "【", "》": "《", "”": "“", "’": "‘"}
+_REF_TRAILING_NOISE_CHARS = ".,;:)]}\"'`\u3002\uff0c\uff1b\uff1a\u3001\uff09\u3011\u300b\u300d\u300f\u201d\u2019"
+_REF_CLOSING_DELIMITERS = {")": "(", "]": "[", "}": "{", "）": "（", "】": "【", "》": "《", "」": "「", "』": "『", "”": "“", "’": "‘"}
 
 # Content-block and structured-key refs are trusted only in these shapes.
 # `data:`/`blob:` URIs can carry arbitrarily large embedded payloads (MCP

backend/tests/test_artifact_integration.py

@@ -5,16 +5,20 @@ context -> model references the handle in a later tool call -> resolution
 rewrites it to the real reference at the tool-call boundary.
 """
 
+import json
 import re
+from pathlib import PurePosixPath
+from types import SimpleNamespace
 from typing import Any
 
 import httpx
 import pytest
 from _agent_e2e_helpers import FakeToolCallingModel
 from langchain.agents import create_agent
-from langchain_core.messages import AIMessage, HumanMessage
+from langchain_core.messages import AIMessage, HumanMessage, ToolMessage
 from langchain_core.tools import tool
 from langgraph.checkpoint.memory import InMemorySaver
+from langgraph.prebuilt.tool_node import ToolCallRequest
 
 from deerflow.agents.middlewares.artifact_capture_middleware import ArtifactCaptureMiddleware
 from deerflow.agents.middlewares.artifact_resolution_middleware import ArtifactResolutionMiddleware
@@ -223,3 +227,36 @@ def test_glob_directory_survives_capture_checkpoint_and_resolved_listing(tmp_pat
     assert found["consumed_by"] == ["call_ls"]
     persisted = next(entry for entry in agent.get_state(config).values["tool_artifacts"] if entry["tool_call_id"] == "call_glob")
     assert persisted["real_ref"] == real_ref
+
+
+@pytest.mark.parametrize(("opener", "closer"), [("「", "」"), ("『", "』")])
+def test_corner_quoted_artifact_resolves_after_source_message_is_gone(tmp_path, opener, closer):
+    filename = "季度 报告.txt"
+    (tmp_path / filename).write_text("report contents", encoding="utf-8")
+    real_ref = f"/mnt/user-data/outputs/{filename}"
+    runtime = SimpleNamespace(context={"thread_id": THREAD_ID})
+    source = ToolMessage(id="corner-report", content=f"已生成{opener}{real_ref}{closer}。", tool_call_id="call_make", name="make_file")
+    update = ArtifactCaptureMiddleware().before_model({"messages": [source]}, runtime)
+    assert update is not None
+    # Restore only the persisted registry: the tool result may have been compacted away.
+    registry = json.loads(json.dumps(update["tool_artifacts"]))
+    assert len(registry) == 1
+    request = ToolCallRequest(
+        tool_call={"name": "read_file", "args": {"path": registry[0]["handle"]}, "id": "call_read", "type": "tool_call"},
+        tool=None,
+        state={"messages": [], "tool_artifacts": registry},
+        runtime=runtime,
+    )
+    reads = []
+
+    def read_report(resolved):
+        path = resolved.tool_call["args"]["path"]
+        reads.append(path)
+        relative = PurePosixPath(path).relative_to("/mnt/user-data/outputs")
+        return ToolMessage(content=(tmp_path / str(relative)).read_text(encoding="utf-8"), tool_call_id="call_read")
+
+    result = ArtifactResolutionMiddleware().wrap_tool_call(request, read_report)
+
+    assert reads == [real_ref]
+    assert result.content == "report contents"
+    assert registry[0]["display_name"] == filename

backend/tests/test_artifact_registry.py

@@ -813,7 +813,7 @@ def test_escaped_sandbox_path_requires_a_matching_escaped_closer():
 
 
 _LINE_BREAKS = ["\n", "\r", "\r\n", "\v", "\f", "\x1c", "\x1d", "\x1e", "\x85", "\u2028", "\u2029"]
-_CJK_QUOTE_PAIRS = [("“", "”"), ("‘", "’")]
+_CJK_QUOTE_PAIRS = [("“", "”"), ("‘", "’"), ("「", "」"), ("『", "』")]
 
 
 @pytest.mark.parametrize("separator", [*_LINE_BREAKS, "\t"])
@@ -844,7 +844,7 @@ def test_escaped_quote_wrappers_do_not_append_backslashes(delimiter, escape_coun
     assert refs[0]["display"] == display
 
 
-@pytest.mark.parametrize(("opener", "closer"), [("“", "”"), ("‘", "’")])
+@pytest.mark.parametrize(("opener", "closer"), _CJK_QUOTE_PAIRS)
 def test_cjk_quote_pairs_keep_the_entire_sandbox_path(opener, closer):
     path = "/mnt/user-data/outputs/输出 报告（最终）.pdf"
 
@@ -854,11 +854,18 @@ def test_cjk_quote_pairs_keep_the_entire_sandbox_path(opener, closer):
     assert refs[0]["display"] == "输出 报告（最终）.pdf"
 
 
-@pytest.mark.parametrize("opener", ["“", "‘"])
+@pytest.mark.parametrize("opener", [opener for opener, _ in _CJK_QUOTE_PAIRS])
 def test_unterminated_cjk_quotes_do_not_register_prefixes(opener):
     assert _detect_refs_in_text(f"已创建 {opener}/mnt/user-data/outputs/输出 报告.pdf") == []
 
 
+@pytest.mark.parametrize("closer", [closer for _, closer in _CJK_QUOTE_PAIRS[-2:]])
+def test_a_bare_path_drops_an_unpaired_corner_closer(closer):
+    path = "/mnt/user-data/outputs/报告.pdf"
+
+    assert [ref["ref"] for ref in _detect_refs_in_text(f"已生成 {path}{closer}，")] == [path]
+
+
 def test_even_backslashes_before_a_quote_preserve_literal_path_backslashes():
     prefix = "\\" * 2
     path = "/mnt/user-data/outputs/a b\\"
@@ -866,7 +873,7 @@ def test_even_backslashes_before_a_quote_preserve_literal_path_backslashes():
     assert [ref["ref"] for ref in _detect_refs_in_text(f'{prefix}"{path}"')] == [path]
 
 
-@pytest.mark.parametrize(("opener", "closer"), [('\\"', '\\"'), ("“", "”"), ("‘", "’")])
+@pytest.mark.parametrize(("opener", "closer"), [('\\"', '\\"'), *_CJK_QUOTE_PAIRS])
 @pytest.mark.parametrize("text_block", [False, True])
 def test_review_quote_wrappers_deduplicate_structured_refs_and_keep_order(opener, closer, text_block):
     first = "/mnt/user-data/outputs/输出 报告.pdf"