forced-feedback-loop

Enforce budget-bounded feedback loops with structured work-logs and a searchable knowledge base.

18|Updated Mar 3, 2026
One-click install
npx skills add https://github.com/Eunho-J/timework-skill --skill forced-feedback-loop
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: forced-feedback-loop
Source: https://github.com/Eunho-J/timework-skill/tree/main
Command: npx skills add https://github.com/Eunho-J/timework-skill --skill forced-feedback-loop

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Stalled, one-pass work is common; this Skill enforces continuous self-improvement loops until the allocated budget is exhausted.

Core Features & Use Cases

  • Enforces time- or loop-based budgets to sustain deep work across tasks (research, code, design, document, project, analysis).
  • Records every loop in a structured work-log and updates a KB with convergent/divergent proposals.
  • Uses a search sub-agent to surface relevant prior notes and maintain traceability of decisions.

Quick Start

Begin a guided, time- or loop-limited session on your chosen task and log each iteration in work-log.

Frequently Asked Questions about forced-feedback-loop

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

FAQPage Schema
How do I prevent iterative work from stalling before a task is finished?

A structured work-log captures every iteration, while a knowledge base records convergent and divergent proposals. A dedicated search sub-agent surfaces prior notes to maintain decision traceability, linking outcomes directly to evidence.

How do I set up a time-limited feedback loop for design iterations?

Yes, loop-count-based termination is supported alongside time-based modes. You can configure either or both termination modes to control iterative improvements, ensuring work stops precisely when the allocated loop budget is exhausted.

How does a knowledge base track divergent and convergent proposals during analysis?

The knowledge base tracks divergent and convergent proposals by recording them during each feedback loop iteration. A dedicated search sub-agent queries this knowledge base to surface relevant prior notes and trace decisions back to their evidence.

Can I use loop-count-based termination for coding iterations instead of a time limit?

Yes, loop-count-based termination is supported alongside time-based modes. You can configure either or both termination modes to control iterative improvements, ensuring work stops precisely when the allocated loop budget is exhausted.