wanwu/deer-flow
1
0
Fork 0
mirror of https://gitee.com/wanwujie/deer-flow synced 2026-04-19 04:14:46 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix: Long thinking but with empty content (#12)

Browse Source
This commit is contained in:
DanielWalnut
2026-01-18 14:21:40 +08:00
committed by GitHub
parent 5f4c58aa82
commit 6f97dde5d1
3 changed files with 8 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
backend/debug.py
View File

@@ -32,7 +32,7 @@ async def main():
"thread_id": "debug-thread-001", "thread_id": "debug-thread-001",
"thinking_enabled": True, "thinking_enabled": True,
# Uncomment to use a specific model # Uncomment to use a specific model
"model_name": "doubao-seed-1.8", "model_name": "deepseek-v3.2",
} }
} }

5
backend/docs/TODO.md
View File

@@ -2,10 +2,11 @@
## Features ## Features
[ ] Launch the sandbox only after the first file system or bash tool is called [x] Launch the sandbox only after the first file system or bash tool is called
[ ] Pooling the sandbox resources to reduce the number of sandbox containers [ ] Pooling the sandbox resources to reduce the number of sandbox containers
[ ] Add Clarification Process for the whole process
## Issues ## Issues
[x] Make sure that no duplicated files in `state.artifacts` [x] Make sure that no duplicated files in `state.artifacts`
[ ] Long thinking but with empty content (answer inside thinking process) [x] Long thinking but with empty content (answer inside thinking process)

11
backend/src/agents/lead_agent/prompt.py
View File

@@ -8,8 +8,10 @@ You are DeerFlow 2.0, an open-source super agent.
</role> </role>
<thinking_style> <thinking_style>
- Think concisely - Think concisely and strategically
- Never write down your full final answer or report in thinking process, but only outline - Never write down your full final answer or report in thinking process, but only outline
- CRITICAL: After thinking, you MUST provide your actual response to the user. Thinking is for planning, the response is for delivery.
- Your response must contain the actual answer, not just a reference to what you thought about
</thinking_style> </thinking_style>
<skill_system> <skill_system>
@@ -45,12 +47,6 @@ All temporary work happens in `/mnt/user-data/workspace`. Final deliverables mus
- Action-Oriented: Focus on delivering results, not explaining processes - Action-Oriented: Focus on delivering results, not explaining processes
</response_style> </response_style>
<memory_and_context>
- Software engineer and prompt engineer at Bytedance Shanghai
- Tech stack: TypeScript, Next.js, Tailwind v4, Shadcn, Python
- Working on AIGC with Gemini Nano Banana
</memory_and_context>
<critical_reminders> <critical_reminders>
- Skill First: Always load the relevant skill before starting **complex** tasks. - Skill First: Always load the relevant skill before starting **complex** tasks.
- Progressive Loading: Load resources incrementally as referenced in skills - Progressive Loading: Load resources incrementally as referenced in skills
@@ -58,6 +54,7 @@ All temporary work happens in `/mnt/user-data/workspace`. Final deliverables mus
- Clarity: Be direct and helpful, avoid unnecessary meta-commentary - Clarity: Be direct and helpful, avoid unnecessary meta-commentary
- Multi-task: Better utilize parallel tool calling to call multiple tools at one time for better performance - Multi-task: Better utilize parallel tool calling to call multiple tools at one time for better performance
- Language Consistency: Keep using the same language as user's - Language Consistency: Keep using the same language as user's
- Always Respond: Your thinking is internal. You MUST always provide a visible response to the user after thinking.
</critical_reminders> </critical_reminders>
""" """