skills.rest
felipelobomotta-blip
felipelobomotta-blipCOMMUNITY
Explore all skills by this author →
GitHub

quality-gate

Orchestrate evaluate-fix-reevaluate loops enforcing editorial thresholds for book chapters.

92|28|Updated Apr 13, 2026
Search Tags:#adversarial review#llm orchestration#editorial feedback#chapter evaluation#revision loop#quality gating#stateful pipelines
One-click install
npx skills add https://github.com/felipelobomotta-blip/book-genesis-v4 --skill quality-gate-felipelobomotta-blip
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: quality-gate
Source: https://github.com/felipelobomotta-blip/book-genesis-v4/tree/main/skills/deprecated/quality-gate
Command: npx skills add https://github.com/felipelobomotta-blip/book-genesis-v4 --skill quality-gate-felipelobomotta-blip

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill removes manual babysitting from an evaluate–fix–re-evaluate workflow by automatically enforcing chapter quality thresholds and dispatching targeted revisions until a chapter passes.

Core Features & Use Cases

  • Automated evaluate–fix–re-evaluate loop: Orchestrates evaluation by sending chapters to /beta-reader, then routes revision requests to /book-editor or /narrative-foundation based on gate outcomes.
  • Configurable hard gates with near-miss handling: Enforces Genesis Score floors, anti-AI pattern limits, Pattern #11 limits, em-dash caps, and requires the evaluator’s Casual Reader verdict to be "keep reading".
  • Actionable feedback synthesis: Translates evaluator findings into ranked, editor-ready instructions with paragraph-level specificity and verification steps.

Quick Start

Run the quality-gate skill to evaluate a chapter, enforce thresholds, dispatch revisions to the correct downstream agent, then re-evaluate until it passes or escalates.

Frequently Asked Questions about quality-gate

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

FAQPage Schema
How do I automate the evaluate-fix-reevaluate loop for long-form chapter revisions?▼

Automated chapter revision orchestrates evaluation by dispatching to beta-reader agents, routing targeted fixes to book-editor or narrative-foundation agents, and re-evaluating until configurable editorial thresholds pass. This removes manual babysitting from multi-agent long-form writing pipelines.

What quality gates and anti-AI thresholds should I enforce for Llm-generated book chapters?▼

Chapter quality gates should enforce Genesis Score floors, anti-AI pattern limits, Pattern #11 limits, em-dash caps, and require the evaluator's Casual Reader verdict to be "keep reading" before advancing the stateful pipeline.

How does editorial feedback synthesis work for chapter evaluation in stateful pipelines?▼

Editorial feedback synthesis translates evaluator findings into ranked, editor-ready revision instructions with paragraph-level specificity and verification steps, routing them directly to downstream agents for blind re-evaluation.

Do I need state files like STATE.yaml and outline.md to run a chapter quality gate?▼

Yes, running an automated chapter quality gate requires reading STATE.yaml, foundation.md, outline.md, and voice-bank targets to enforce editorial thresholds and dispatch revisions correctly within the writing pipeline.

Can I configure near-miss handling when a chapter quality gate fails by a small margin?▼

Yes, the quality gate supports configurable hard gates with near-miss handling, allowing chapters that narrowly miss scoring or anti-AI thresholds to route targeted revisions instead of failing outright.

What is the best way to handle adversarial review for multi-agent book writing pipelines?▼

Adversarial review for multi-agent book writing is best handled by an automated quality loop that dispatches chapters to beta-reader evaluators, enforces anti-AI and scoring thresholds, and routes feedback to editors until passing.

Related Skills

Explore More Content & Communication Skills→
Lev0n82Lev0n82
1

natural-dialogue-techniques

Define per-character voice profiles and enforce subtext in narrative dialogue.

Community
Intermediate
cdeistopenedcdeistopened
2

book-quality-loop

Enforce a five-judge quality gate before book chapter publication.

Community
Advanced
TbK0ngTbK0ng

parallel

Coordinate Codex-powered agents across planning, dispatch, and execution phases.

Community
Advanced
roanbrasilroanbrasil
5

llm-evals

Implement LLM evaluation patterns with LangSmith, RAGAS, and PromptFoo.

Community
Advanced