SWE-agent Tutorial: Autonomous Repository Repair and Benchmark-Driven Engineering

Learn how to use SWE-agent/SWE-agent for autonomous software engineering workflows, from single-issue runs to benchmark and research-grade evaluation.

Why This Track Matters

SWE-agent remains one of the strongest open-source references for autonomous software engineering loops, especially for repo-scale issue resolution and SWE-bench workflows.

This track focuses on:

• running SWE-agent quickly on real repository issues
• understanding YAML-driven orchestration and tool contracts
• configuring model, environment, and execution controls
• operating evaluation and production workflows responsibly

Current Snapshot (auto-updated)

• repository: SWE-agent/SWE-agent
• stars: about 19.2k
• latest release: v1.1.0 (published 2025-05-22)

Mental Model

flowchart LR
    A[Issue or Task] --> B[Prompt and Config YAML]
    B --> C[Agent Reasoning Loop]
    C --> D[Tooling and Environment Actions]
    D --> E[Patch and Validation Output]
    E --> F[Benchmark or PR Workflow]

Chapter Guide

Chapter	Key Question	Outcome
01 - Getting Started	How do I run SWE-agent on a first task?	Working baseline
02 - Core Architecture and YAML Configuration	How is SWE-agent structured and configured?	Architecture clarity
03 - CLI Workflows and Usage Modes	How do I run single-task and batch flows effectively?	Execution strategy
04 - Tooling, Environments, and Model Strategy	How do I tune tools, envs, and models safely?	Better runtime control
05 - Benchmarking and Evaluation Practices	How do I evaluate quality and regressions?	Strong evaluation loop
06 - Offensive Security Mode and Specialized Workloads	How do specialized modes like EnIGMA fit in?	Workload scoping
07 - Development and Contribution Workflow	How do I contribute effectively to SWE-agent?	Contributor readiness
08 - Production Operations and Governance	How do teams operate SWE-agent safely over time?	Governance runbook

What You Will Learn

• how to configure and run SWE-agent across real-world coding tasks
• how to reason about YAML config surfaces and tool orchestration
• how to evaluate outcomes with benchmark-oriented discipline
• how to govern autonomous coding workflows in team environments

Source References

• SWE-agent Repository
• SWE-agent README
• SWE-agent Docs
• Hello World Usage
• Batch Mode Usage
• Development Contribution Docs

Related Tutorials

• Open SWE Tutorial
• OpenHands Tutorial
• LangGraph Tutorial
• Cline Tutorial

---

Start with Chapter 1: Getting Started.

Navigation & Backlinks

• Start Here: Chapter 1: Getting Started
• Back to Main Catalog
• Browse A-Z Tutorial Directory
• Search by Intent
• Explore Category Hubs

Full Chapter Map

1. Chapter 1: Getting Started
2. Chapter 2: Core Architecture and YAML Configuration
3. Chapter 3: CLI Workflows and Usage Modes
4. Chapter 4: Tooling, Environments, and Model Strategy
5. Chapter 5: Benchmarking and Evaluation Practices
6. Chapter 6: Offensive Security Mode and Specialized Workloads
7. Chapter 7: Development and Contribution Workflow
8. Chapter 8: Production Operations and Governance

Generated by AI Codebase Knowledge Builder