loop-eng
NewLoop engineering with a hard gate: npx loop-engineering install adds a SKILL.md for Claude Code, Codex, Cursor, Windsurf, and Antigravity that refuses to start a loop until the goal has a real verifier, an iteration cap, and a no-progress exit. Includes loop-verify, loop-run, loop-audit, loop-cost.
Summary
This skill enforces safe loop engineering by requiring a verifiable goal, an iteration cap, and a no-progress exit before any loop begins.
- It provides commands to verify, run, audit, and cost-analyze loops, preventing infinite or wasteful iterations in Claude Code, Codex, Cursor, Windsurf, and Antigravity.
Install & Usage
mkdir -p .claude/skillsmkdir -p .claude/skills && curl -o .claude/skills/loop-eng.md https://raw.githubusercontent.com/rithvikshettyy/loop-eng/main/SKILL.md/loop-engUse Cases
Usage Examples
/loop-eng loop-run --goal 'All unit tests pass' --verifier 'npm test' --max-iterations 5 --exit-on-no-progress
/loop-eng loop-verify --goal 'Reduce bundle size by 10%' --verifier 'npm run build && check-size'
/loop-eng loop-cost --goal 'Add pagination to API' --max-iterations 20
Security Audits
Frequently Asked Questions
What is loop-eng?
This skill enforces safe loop engineering by requiring a verifiable goal, an iteration cap, and a no-progress exit before any loop begins. It provides commands to verify, run, audit, and cost-analyze loops, preventing infinite or wasteful iterations in Claude Code, Codex, Cursor, Windsurf, and Antigravity.
How to install loop-eng?
To install loop-eng: create the skills directory (mkdir -p .claude/skills), then run: mkdir -p .claude/skills && curl -o .claude/skills/loop-eng.md https://raw.githubusercontent.com/rithvikshettyy/loop-eng/main/SKILL.md. Finally, /loop-eng in Claude Code.
What is loop-eng best for?
loop-eng is a skill categorized under General. Created by rithvikshettyy.
What can I use loop-eng for?
loop-eng is useful for: When you need to run a code refactoring loop that must stop after 10 iterations or when no further improvements are detected.; When you want to audit a loop's history to see how many iterations were executed and what changes were made.; When you need to estimate the cost (in tokens or time) of a proposed loop before running it.; When you want to ensure a bug-fixing loop has a concrete test that must pass before the loop stops.; When you need to enforce a hard iteration cap on a code generation loop to avoid runaway costs.; When you want to verify that a loop's goal is measurable and has a defined exit condition before starting..