loop-agent

Orchestrate workflow agents in iterative refinement loops with configurable termination modes.

11|Updated Nov 5, 2025
One-click install
npx skills add https://github.com/d-o-hub/rust-self-learning-memory --skill loop-agent
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: loop-agent
Source: https://github.com/d-o-hub/rust-self-learning-memory/tree/main/.claude/skills/loop-agent
Command: npx skills add https://github.com/d-o-hub/rust-self-learning-memory --skill loop-agent

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill enables systematic iterative workflows with convergence detection, progress tracking, and intelligent termination to progressively improve outputs and quality.

Core Features & Use Cases

  • Iteration Planning: configure fixed, criteria-based, convergence-based, or hybrid termination.
  • Progress Tracking: monitor iteration counts, success criteria, and improvements.
  • Execution Patterns: define cycles for refinement in code, tests, performance, or documentation.

Quick Start

Define a loop plan with max iterations and a success criterion; run iterations until the target is met or max iterations reached.

Frequently Asked Questions about loop-agent

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

FAQPage Schema
How do I set up iterative refinement loops for code or testing workflows?

Iterative refinement loops orchestrate repeated cycles of improvement by defining a loop plan with max iterations and a success criterion. Configure termination modes—fixed iteration count, criteria-based checks, convergence detection, or hybrid—then run cycles until your target quality is met or max iterations reached, with progress tracking throughout.

What's the difference between fixed iteration, criteria-based, and convergence termination modes?

Termination modes control when loops stop: fixed iteration runs a set number of cycles regardless of outcome; criteria-based halts when specific success conditions are met; convergence stops when improvements plateau; hybrid combines multiple modes. Choose based on whether your goal is predictable cycles, quality thresholds, or detecting diminishing returns.

Can I use iterative refinement for performance optimization and documentation workflows?

Yes, iterative refinement applies to code refinement, testing validation, performance tuning, and documentation analysis. Each iteration applies fix–validate cycles, monitors progress against success criteria, and detects convergence, enabling systematic quality improvement across any workflow requiring repeated cycles and measurable advancement.

How do I track progress and know when to stop iterating?

Progress tracking monitors iteration counts, success criteria achievement, and improvement metrics across cycles. Convergence checks detect when further iterations yield diminishing returns. Combined with your chosen termination mode—criteria-based thresholds, fixed limits, or convergence signals—you gain visibility into when quality targets are reached.

What agents or components do I need to orchestrate iterations?

Define workflow agents to execute within each iteration cycle; the loop plan orchestrates their sequencing per iteration. No predefined dependencies or components are required—you configure agents specific to your refinement task, whether code analysis, test validation, or performance measurement, and the loop manages their execution until termination.

When should I avoid iterative refinement loops?

Avoid iterative refinement when tasks require single-pass execution with no feedback loops, outcomes are unpredictable across cycles, or convergence is impossible to define. Iterative refinement excels at progressive improvement but demands measurable success criteria and agent cycles capable of incremental advancement toward quality targets.