# KodeKloud: Beta Courses Loop Engineering

## Loop Engineering

### What Is a Loop

- [Introduction](https://notes.kodekloud.com/docs/Loop-Engineering/What-Is-a-Loop/Introduction/page.md): Introduction to loop engineering for autonomous coding agents, teaching the Try Test Fix Save cycle, scorekeeping, and components like automations, worktrees, skills, connectors, sub agents, and memory
- [What is Loop Engineering](https://notes.kodekloud.com/docs/Loop-Engineering/What-Is-a-Loop/What-is-Loop-Engineering/page.md): Explains loop engineering where AI coding agents repeatedly try, test, fix, and save small changes to autonomously build reliable software using clear goals and measurable tests.
- [The Scorekeeper](https://notes.kodekloud.com/docs/Loop-Engineering/What-Is-a-Loop/The-Scorekeeper/page.md): Explains the importance of an objective, automated, fast, and tamperproof scorekeeper to measure and guide iterative loops and prevent gaming
- [The Six Components of Loop Engineering](https://notes.kodekloud.com/docs/Loop-Engineering/What-Is-a-Loop/The-Six-Components-of-Loop-Engineering/page.md): Summarizes six components of loop engineering—automation, worktrees, skills, connectors, subagents, and memory—for building reliable autonomous loops that iteratively improve systems

### Components

- [Automations](https://notes.kodekloud.com/docs/Loop-Engineering/Components/Automations/page.md): Explains automation in loop engineering by repeating tasks, evaluating results, applying incremental fixes, saving progress, using timer or run-until-done loops, and enforcing guardrails for safe improvement.
- [Worktrees](https://notes.kodekloud.com/docs/Loop-Engineering/Components/Worktrees/page.md): Explains Git worktrees and how to create, list, remove, and prune separate working copies for safe experiments and CI.
- [Skills](https://notes.kodekloud.com/docs/Loop-Engineering/Components/Skills/page.md): How to create and use compact skill.md files that package focused, testable steps for automation agents to load, ensuring modular, predictable, and reusable behavior in an automation loop
- [Connectors Plugins](https://notes.kodekloud.com/docs/Loop-Engineering/Components/Connectors-Plugins/page.md): Describes connectors and plugins that link the Loop to external tools, MCP standard, and security best practices for permissions and secret management
- [Sub agents](https://notes.kodekloud.com/docs/Loop-Engineering/Components/Sub-agents/page.md): Using isolated sub-agents to offload single focused tasks from a main loop, improving speed, safety, and parallelism while keeping workspace clean and enabling maker-and-checker validation.
- [Memory](https://notes.kodekloud.com/docs/Loop-Engineering/Components/Memory/page.md): Explains using small persistent memory in looped systems with three tiers and five concise notes to resume work across sessions, avoid repeats, and prevent context drift.

### Tie It All Together

- [The Whole Loop in Motion](https://notes.kodekloud.com/docs/Loop-Engineering/Tie-It-All-Together/The-Whole-Loop-in-Motion/page.md): Explains an automated iterative loop using goals, scorekeepers, memory, worktrees, skills, and helpers to make small validated changes until a goal is achieved, with human review for risky decisions
- [Designing Your Own Loop](https://notes.kodekloud.com/docs/Loop-Engineering/Tie-It-All-Together/Designing-Your-Own-Loop/page.md): Design safe, incremental autonomous loops by setting narrow goals, creating objective scorekeepers, getting one saved successful run, and adding components only to solve observed failures.

### Conclusion

- [Wrap Up and Conclusion](https://notes.kodekloud.com/docs/Loop-Engineering/Conclusion/Wrap-Up-and-Conclusion/page.md): Guide to building AI coding agents using a try‑test‑fix‑save loop, prioritizing goal and scorekeeper, adding components only as needed for reliable automation.