Ironloop

Verification-first software engineering harness with 5 layers (spec → gen → test → sim → pentest). Use when generating, reviewing, or planning code with AI agents, especially for Rust projects, new feature work, legacy migrations, or any task where code correctness must be guaranteed. Activates on phrases like "build a", "implement", "refactor", "review this code", "migrate from", "add tests for", or when the user describes a coding task with correctness or security requirements.

10 starsProductivity & Notes

Install

skills CLI (Claude Code, Cursor, Codex and more)

npx skills add edouard-claude/ironloop --skill ironloop

Manual install

git clone --depth 1 https://github.com/edouard-claude/ironloop.git /tmp/ironloop
mkdir -p ~/.claude/skills
cp -r /tmp/ironloop/skills/engineering/ironloop ~/.claude/skills/ironloop

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# IRONLOOP v1.7

You are IRONLOOP, a software engineering system built on one principle:
**code is disposable, the harness is permanent.**

You do not write code. You build a harness *around* code: five layers of
verification that make the code irrelevant. When the harness passes, the
code is correct. When it doesn't, the code gets regenerated until it does.

## Locked Vocabulary

Use these terms exactly. Do not substitute synonyms; consistent language
is the point.

| Term | Definition | Never say |
|------|-----------|-----------|
| **layer** | One of the 5 verification layers | "step", "phase", "stage" |
| **harness** | The entire 5-layer harness around the code | "pipeline", "framework", "workflow" |
| **loop** | A feedback cycle within a layer | "cycle", "iteration" |
| **generation** | Code produced by the agent | "coding", "implementation" |
| **verification** | Any check that the code is correct | "validation", "checking" |
| **specification** | The human-written contract (Layer 1 output) | "spec", "requirements" |

## Leading Words

These tokens anchor specific behaviors. Use them, don't paraphrase them.

- **red-capable**: a test that CAN fail. It asserts the exact symptom, not
  just "runs without error". A test that never goes red proves nothing.
  Every test in Layer 3 must be red-capable, and Layer 3 proves it with
  mutation testing: a surviving mutant is a test that is not red-capable.
- **red**: a test is red when it *compiles* and *fails on an assertion*.

Read the full skill on GitHub

Details

Skill name: ironloop
Repository: edouard-claude/ironloop
Path: skills/engineering/ironloop/SKILL.md
License: MIT
Repo updated: 11 September 2026
Listed in: karanb192/awesome-claude-skills

FAQ

What does the Ironloop skill do?

How do I install the Ironloop skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add edouard-claude/ironloop --skill ironloop. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Ironloop popular?

It lives in edouard-claude/ironloop, which has 10 GitHub stars and 0 forks, and is included in 1 curated skill list.

Related skills

Webman

zjkal/webman-skill

Expert skill for the webman framework (a long-lived, in-memory PHP framework based on workerman). Covers routing, controllers, middleware, database/Redis, custom processes, timers, coroutines (v2), plugin development, and guarding against memory leaks and cross-request state pollution under the resident process model. Use when the project has start.php, config/process.php, or support/bootstrap.php, or when the user mentions webman, workerman, 常驻内存 PHP, 协程 PHP, or long-lived/in-memory PHP.

Productivity & Notes

Specmint Core

ngvoicu/specmint-core

Persistent spec management for AI coding workflows. Use this skill when the user explicitly mentions specs, forging, or structured planning: says "forge", "forge a spec", "write a spec for X", "create a spec", "plan X as a spec", "resume", "what was I working on", "spec list/status/pause/switch/activate", "implement the spec", "implement phase N", "implement all phases", "generate openapi", or exits plan mode (offer to save as a spec). Also trigger when a `.specs/` directory exists at session start. Do NOT trigger on general feature requests, coding tasks, or questions that don't mention specs

Productivity & Notes

Feature Marker

Viniciuscarvalho/Feature-marker

Skill-first run-through feature workflow for Claude, Codex, and Gemini. Use when the user asks to implement a feature through PRD, TechSpec, Tasks, verification, local commit, and branch handoff. The npm package installs skill files only; the LLM skill performs the workflow.

Productivity & Notes

Solve Math Rigorously

googlarz/math-skill

World-class solver-first math skill for reading, mapping, solving, and checking nontrivial math tasks. Use when the user asks to solve an equation or system, find all solutions, prove a claim, optimize a function, work through a derivation step by step, check whether an answer is correct, explain what a worksheet or whiteboard problem says, or read math from an uploaded photo, screenshot, PDF, textbook page, handwritten note, or diagram before solving it carefully and verifying the result.

Productivity & Notes

Patton Protocol

vladimirrott/patton-protocol

Coordinate bounded coding-agent missions when independent delegation improves the task.

Productivity & Notes

Privacy Mask

fullstackcrew-alpha/privacy-mask

Mask, redact, anonymize and censor sensitive information (PII) in screenshots and images — phone numbers, emails, IDs, API keys, crypto wallets, credit cards, passwords, and more. Uses OCR (Tesseract + RapidOCR) with 47 regex rules and optional NER (GLiNER) to detect private data and applies blur/fill redaction overlays. Supports GDPR and compliance workflows, secret detection, and data loss prevention (DLP). All processing runs 100% locally and offline — no data leaves your machine.

Productivity & Notes

Browse the most popular Claude skills, all productivity & notes skills, or write your own with the skills generator.