[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-80132":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":8,"htmlUrl":8,"language":9,"languages":8,"totalLinesOfCode":8,"stars":10,"forks":11,"watchers":12,"openIssues":11,"contributorsCount":11,"subscribersCount":11,"size":11,"stars1d":11,"stars7d":11,"stars30d":13,"stars90d":11,"forks30d":11,"starsTrendScore":11,"compositeScore":11,"rankGlobal":8,"rankLanguage":8,"license":8,"archived":14,"fork":14,"defaultBranch":15,"hasWiki":16,"hasPages":14,"topics":17,"createdAt":8,"pushedAt":8,"updatedAt":18,"readmeContent":19,"aiSummary":20,"trendingCount":11,"starSnapshotCount":11,"syncStatus":13,"lastSyncTime":21,"discoverSource":22},80132,"AST-Guided-LLM-Code-Review_System","rahuldr556\u002FAST-Guided-LLM-Code-Review_System","rahuldr556",null,"Python",56,0,54,2,false,"main",true,[],"2026-06-12 02:03:58","# AST-Guided LLM Code Review Framework v2\n\nA 5-stage pipeline for automated Python bug detection combining AST analysis with LLM reasoning.\n\n## Results\n\n### 80-Case Synthetic Benchmark (Simulation Mode)\n- **Precision: 100%** | **Recall: 94.3%** | **F1: 97.1%**\n\n### 12-Case Curated Real-World (Simulation Mode)\n- **Precision: 100%** | **Recall: 92%** | **F1: 95.8%**\n\n### 90-Case Real-World (GPT-4)\n- **Precision: 100%** | **Recall: 88%** | **F1: 93.6%**\n\n## Quick Start\n\n```bash\npip install openai --break-system-packages\n\n# Simulation mode (free)\npython -m src.main examples\u002Fbuggy_ml_pipeline.py\n\n# GPT-4 mode\nexport OPENAI_API_KEY=\"sk-...\"\npython -m src.main examples\u002Fbuggy_ml_pipeline.py --use-llm\n\n# Benchmarks\npython -m src.paper_eval                               # 80-case synthetic\npython -m data.run_real_eval --curated-only             # Real-world\npython -m data.run_real_eval --curated-only --use-llm   # Real-world + GPT-4\n```\n\n## Architecture\n\n```\nSource Code → [Parse] → [Detect] → [Slice] → [Review] → [Aggregate] → Findings\n              Stage 1    Stage 2    Stage 3    Stage 4     Stage 5\n```\n\n**Novel Contributions:** Adaptive Slicing, Confidence Calibration (Platt scaling), Cross-Detector Correlation\n\n**6 Bug Categories:** None-deref, Division-by-zero, Index-OOB, Data Leakage, Shape Mismatch, Resource Leak\n\n**39 unit tests, 92 benchmark cases, 25 Python files, ~8500 LOC**\n","该项目是一个结合抽象语法树（AST）分析与大语言模型推理的五阶段Python代码审查框架，旨在自动检测代码中的错误。其核心功能包括源代码解析、缺陷检测、代码切片、代码审查和结果汇总五个步骤，并引入了自适应切片、置信度校准以及跨检测器相关性等创新技术。该系统特别适用于需要高效准确地识别如空指针引用、除零错误、数组越界访问、数据泄露、形状不匹配及资源泄漏等常见编程错误的场景中。基于公开测试结果显示，在合成与真实案例中均表现出高精度与良好的召回率，证明了其实用价值。","2026-06-11 03:59:21","CREATED_QUERY"]