validation-loop

Orchestrate multi-agent verification loops to validate code changes against quality gates.

41|4|Updated Jan 28, 2026
One-click install
npx skills add https://github.com/etr/groundwork --skill validation-loop
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: validation-loop
Source: https://github.com/etr/groundwork/tree/main/skills/validation-loop
Command: npx skills add https://github.com/etr/groundwork --skill validation-loop

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Manual verification of code changes is error-prone and time-consuming. The Validation Loop orchestrates autonomous multi-agent verification to detect issues, propose fixes, and drive consensus until all agents approve.

Core Features & Use Cases

  • End-to-end orchestration of nine specialized verification agents (code-quality, test-quality, security, spec-alignment, architecture-alignment, code-simplifier, housekeeper, performance, and design-consistency).
  • Autonomous fix-and-retry loop: aggregate findings, delegate fixes to a dedicated fixer agent, and re-run impacted agents until convergence.
  • Detailed iteration logging and governance: track issues, domain spillover, and persist unexecuted findings to a minor_todos.md for follow-up.

Quick Start

Run the Validation Loop on the current project to start autonomous verification across all agents.

Frequently Asked Questions about validation-loop

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate multi-agent code verification for monorepo changes?▼

Automate multi-agent code verification by orchestrating nine specialized agents to check changed files, tests, and specs against quality gates. The system aggregates findings, applies fixes, and iterates until all agents approve.

What is the best way to automate code review and quality checks across a monorepo?▼

Automating code review across a monorepo is best handled by a multi-agent verification loop that evaluates code quality, security, and spec alignment. It delegates fixes to a dedicated agent and re-runs checks until convergence.

How does an autonomous fix-and-retry loop work for code quality validation?▼

An autonomous fix-and-retry loop works by aggregating verification findings from multiple agents, delegating the necessary corrections to a fixer agent, and re-running impacted agents until all quality gates pass.

Can I run automated verification agents for spec alignment and security in a monorepo?▼

Yes, you can run automated verification agents for spec alignment and security in a monorepo. The orchestration loop includes specialized agents for security, architecture, and spec alignment to validate changes comprehensively.

What happens to unexecuted issues or minor findings during automated validation?▼

Unexecuted issues and minor findings during automated validation are logged and persisted to a minor_todos.md file. This ensures that domain spillover and unresolved issues are tracked for follow-up without blocking the main loop.

Does multi-agent code validation require manual intervention to reach consensus?▼

Multi-agent code validation does not require manual intervention to reach consensus. It uses strict guardrails and a fixed agent loop that continuously verifies and fixes code until all nine specialized agents approve the changes.