Agentless
Simplified LLM-based approach to automated software engineering without agents
Agentless
Simplified LLM-based approach to automated software engineering without agents
Agentless is an open-source automated software engineering framework that uses a simplified, two-phase approach—localization followed by repair—to fix software bugs, without complex autonomous agent frameworks. The method locates relevant code using hierarchical search, then generates and validates fixes, achieving strong performance on software engineering benchmarks with lower complexity and cost than full agent-based approaches. Researchers in software engineering automation, teams evaluating lightweight alternatives to autonomous coding agents, and organizations benchmarking automated bug repair approaches use Agentless for its simplicity, reproducibility, and competitive performance.
Key Features
- ✓Two-phase approach
- ✓Code localization
- ✓Bug repair
- ✓SWE-bench performance
- ✓Open-source
Quick Info
- Category
- Code & Development
- Pricing
- Free
More Code & Development Tools
GitHub Copilot
Code & DevelopmentThe AI pair programmer trusted by millions of developers
Cursor
Code & DevelopmentThe code editor built around AI from the ground up
Tabnine
Code & DevelopmentPrivacy-first AI code completion
Codeium
Code & DevelopmentFree AI coding assistant with no usage limits