feedback-loop

Official

Targeted remediation for blocked reviews.

AuthorKiwi-Home
Version1.0.0
Installs0

System Documentation

What problem does it solve?

This Skill defines a repeatable remediation loop to resolve blocking findings from reviews, preventing ad-hoc fixes that miss scope, repeat futile attempts, or leave unresolved issues untracked.

Core Features & Use Cases

  • Controlled Iterations: Tracks remediation attempts with iteration markers that record refs targeted, files changed, commit hashes, and approach summaries for auditable retries.
  • Mode Selection: Supports interactive human-in-the-loop prompts or autonomous execution for unattended remediation runs.
  • Failure & Scope Guardrails: Enforces per-finding retry limits, stuck-finding detection, termination conditions, and strict scope constraints to prevent scope creep.
  • Re-review Safety: Forces comprehensive re-reviews after fixes to catch regressions and newly introduced issues.
  • Use Case: After a /review returns a BLOCK verdict, run targeted remediations against MUST-FIX findings, iterate up to the configured max, and produce a final summary of resolved and stuck findings.

Quick Start

Request a targeted remediation run for the brief with blocking findings and allow the loop to attempt up to the configured number of iterations.

Dependency Matrix

Required Modules

None required

Components

references

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: feedback-loop
Download link: https://github.com/Kiwi-Home/ai-workflow/archive/main.zip#feedback-loop

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 223,000+ vetted skills library on demand.