pre-publish-review

Coordinates multi-agent pre-publish assessment for npm releases.

Updated Jun 11, 2026
One-click install
npx skills add https://github.com/HaydernCenterpoint/Codex-Skills --skill pre-publish-review-hayderncenterpoint
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: pre-publish-review
Source: https://github.com/HaydernCenterpoint/Codex-Skills/tree/main/skills/pre-publish-review
Command: npx skills add https://github.com/HaydernCenterpoint/Codex-Skills --skill pre-publish-review-hayderncenterpoint

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Coordinates a multi-layer pre-publish assessment to ensure npm releases are thoroughly reviewed before publishing. It runs an initial change-detection pass, orchestrates up to 16 agents for per-change analysis, performs a holistic review across the changeset, and outputs a release synthesis verdict to guide safe deployment.

Core Features & Use Cases

  • Unpublished Change Detection: Uses the /get-unpublished-changes flow to identify all diffs since the last release and categorize them by scope.
  • Parallel Per-Change Analysis: Spawns up to 10 ultrabrain agents to validate each change group for correctness, edge cases, and pattern adherence.
  • Holistic Release Review: Invokes a 5-agent holistic review to assess goal compliance, QA execution, code quality, security, and context mining across the full changeset.
  • Release Synthesis: Runs a 1-agent oracle to determine overall release readiness, version bump guidance, and deployment risk.
  • Guided Release Gate: Generates a structured, auditable decision path before every npm publish.

Quick Start

Run the pre-publish-review skill before npm publish to validate your release readiness.

Frequently Asked Questions about pre-publish-review

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

FAQPage Schema
How do I run a pre-publish review for an npm package?

To run a pre-publish review, execute the skill before npm publish to detect unpublished changes, orchestrate up to 16 agents for per-change analysis, and generate a release synthesis verdict for safe deployment.

What is multi-agent code review for npm release readiness?

Multi-agent code review for npm release readiness coordinates up to 16 agents across three layers—per-change deep analysis, holistic review, and release synthesis—to assess risks and determine deployment readiness.

How does agent orchestration work for pre-publish security and QA checks?

Agent orchestration for pre-publish checks spawns up to 10 ultrabrain agents for per-change validation and 5 agents for holistic review, assessing correctness, security, code quality, and goal compliance across the full changeset.

Can I use automated code review to determine npm version bumps?

Yes, automated code review can determine npm version bumps by running a 1-agent oracle that analyzes the synthesized changeset results to provide version bump guidance and evaluate deployment risk.

Do I need to detect unpublished changes before running a holistic code review?

Yes, detecting unpublished changes is the required first step that identifies and categorizes diffs since the last release, providing the structured scope needed to spawn agents for the holistic code review.

What are the limitations of using 16 agents for release gating?

The 16-agent release gating approach enforces strict structured outputs and requires coordinating three distinct analysis layers, meaning it is designed specifically for comprehensive pre-publish assessment rather than quick lightweight checks.