You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
All notable changes to DeepCode will be documented in this file.
4
+
5
+
## [1.0.6-jm] - 2025-10-19
6
+
7
+
### Added
8
+
-**Dynamic Model Limit Detection**: New `utils/model_limits.py` module that automatically detects and adapts to any LLM model's token limits and pricing
9
+
-**Loop Detection System**: `utils/loop_detector.py` prevents infinite loops by detecting repeated tool calls, timeouts, and progress stalls
10
+
-**Progress Tracking**: 8-phase progress tracking (5% → 100%) with file-level progress indicators in both UI and terminal
11
+
-**Abort Mechanism**: "Stop Processing" button in UI with global abort flag for clean process termination
12
+
-**Cache Cleanup Scripts**: `start_clean.bat` and `start_clean.ps1` to clear Python cache before starting
13
+
-**Enhanced Error Display**: Real-time error messages in both UI and terminal with timestamps
14
+
-**File Progress Tracking**: Shows files completed/total with estimated time remaining
15
+
16
+
### Fixed
17
+
-**Critical: False Error Detection**: Fixed overly aggressive error detection that was marking successful operations as failures, causing premature abort and empty file generation
18
+
-**Critical: Empty File Generation**: Files now contain actual code instead of being empty (2-byte files)
19
+
-**Unique Folder Naming**: Each project run now creates `paper_{timestamp}` folders instead of reusing `pdf_output`
20
+
-**PDF Save Location**: PDFs now save to `deepcode_lab/papers/` instead of system temp directory
21
+
-**Duplicate Folder Prevention**: Added session state caching to prevent duplicate folder creation on UI reruns
22
+
-**Token Limit Compliance**: Fixed `max_tokens` to respect model limits dynamically (e.g., gpt-4o-mini's 16,384 token limit)
23
+
-**Empty Plan Detection**: System now fails early with clear error messages when initial plan is empty or invalid
24
+
-**Process Hanging**: Fixed infinite loops and hanging on errors - process now exits cleanly
25
+
-**Token Cost Tracking**: Restored accurate token usage and cost display (was showing $0.0000)
26
+
-**PDF to Markdown Conversion**: Fixed automatic conversion and file location handling
27
+
-**Document Segmentation**: Properly uses configured 50K character threshold from `mcp_agent.config.yaml`
28
+
-**Error Propagation**: Abort mechanism now properly stops process after 10 consecutive real errors
29
+
30
+
### Changed
31
+
-**Model-Aware Token Management**: Token limits now adapt automatically based on configured model instead of hardcoded values
32
+
-**Cost Calculation**: Dynamic pricing based on actual model rates (OpenAI, Anthropic)
33
+
-**Retry Logic**: Token limits for retries now respect model maximum (87.5% → 95% → 98% of max)
34
+
-**Segmentation Workflow**: Better integration with code implementation phase
35
+
-**Error Handling**: Enhanced error propagation - errors no longer reported as "success"
36
+
-**UI Display**: Shows project folder name after PDF conversion for better visibility
37
+
-**Terminal Logging**: Added timestamps to all progress messages
38
+
39
+
### Technical Improvements
40
+
- Added document-segmentation server to code implementation workflow for better token management
41
+
- Improved error handling in agent orchestration engine with proper cleanup
42
+
- Enhanced subprocess handling on Windows (hide console windows, prevent hanging)
43
+
- Better LibreOffice detection on Windows using direct path checking
44
+
- Fixed input data format consistency (JSON with `paper_path` key)
45
+
- Added comprehensive logging throughout the pipeline
46
+
- Improved resource cleanup on errors and process termination
47
+
48
+
### Documentation
49
+
- Translated Chinese comments to English in core workflow files
50
+
- Added inline documentation for new utility modules
51
+
- Created startup scripts with clear usage instructions
52
+
53
+
### Breaking Changes
54
+
- None - all changes are backward compatible
55
+
56
+
### Known Issues
57
+
- Terminal may show trailing "Calling Tool..." line after completion (cosmetic display artifact - process completes successfully)
58
+
- Some Chinese comments remain in non-critical files (cli, tools) - translation in progress
0 commit comments