furyhawk/deer-flow
1
0
Fork 0
mirror of https://github.com/bytedance/deer-flow.git synced 2026-05-23 08:25:57 +00:00
Code Issues Packages Projects Releases Wiki Activity

feat(loop-detection): defer warning injection (#2752)

Browse Source 
* fix(loop-detection): defer warn injection to wrap_model_call

The warn branch in LoopDetectionMiddleware injected a HumanMessage
into state from after_model. The tools node had not yet produced
ToolMessage responses to the previous AIMessage(tool_calls=...), so
the new HumanMessage landed *between* the assistant's tool_calls and
their responses. OpenAI/Moonshot reject the next request with
"tool_call_ids did not have response messages" because their
validators require tool_calls to be followed immediately by tool
messages.

Detection now runs in after_model as before, but only enqueues the
warning into a per-thread list. Injection happens in wrap_model_call,
where every prior ToolMessage is already present in request.messages.
The warning is appended at the end as HumanMessage(name="loop_warning")
— pairing intact, AIMessage semantics untouched, no SystemMessage
issues for Anthropic.

Closes #2029, addresses #2255 #2293 #2304 #2511.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

* fix(channels): remove loop warning display filter

* feat(loop-detection): scope pending warnings by run

* docs(loop-detection): update docs

* test(loop-detection): assert deferred warnings are queued

* fix(loop-detection): cap transient warning state

* docs: update docs

* add async awrap_model_call test coverage

* docs(loop-detection): document transient warnings

---------

Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
This commit is contained in:
Nan Gao
2026-05-21 08:36:07 +02:00
committed by GitHub
parent 7ec8d3a6e7
commit dcc6f1e678
7 changed files with 696 additions and 221 deletions
Download Patch File Download Diff File Expand all files Collapse all files
backend/tests/test_channels.py
-31
View File
@@ -372,37 +372,6 @@ class TestExtractResponseText:
# Should return "" (no text in current turn), NOT "Hi there!" from previous turn
assert _extract_response_text(result) == ""
def test_does_not_publish_loop_warning_on_tool_calling_ai_message(self):
"""Loop-detection warning text on a tool-calling AI message is middleware-authored."""
from app.channels.manager import _extract_response_text
result = {
"messages": [
{"type": "human", "content": "search the repo"},
{
"type": "ai",
"content": "[LOOP DETECTED] You are repeating the same tool calls.",
"tool_calls": [{"name": "grep", "args": {"pattern": "TODO"}, "id": "call_1"}],
},
]
}
assert _extract_response_text(result) == ""
def test_preserves_visible_text_when_stripping_loop_warning(self):
from app.channels.manager import _extract_response_text
result = {
"messages": [
{"type": "human", "content": "prepare the report"},
{
"type": "ai",
"content": "Here is the report.\n\n[LOOP DETECTED] You are repeating the same tool calls.",
"tool_calls": [{"name": "present_files", "args": {"filepaths": ["/mnt/user-data/outputs/report.md"]}, "id": "call_1"}],
},
]
}
assert _extract_response_text(result) == "Here is the report."
# ---------------------------------------------------------------------------
# ChannelManager tests