ARB — Publisher Monetization Readiness Platform

English | 中文说明

ARB is a documentation-first audit framework for evaluating whether a website is ready for AdSense review and broader publisher monetization. It combines a benchmark model, a fixed scoring contract, and a 26-skill workflow that covers diagnosis, remediation, verification, monitoring, and recovery.

The framework is built around 6 pillars, a fixed Core 79 review set for baseline checks, optional extension coverage up to the full 105-item benchmark, and hard veto rules so critical risks are never hidden by an average score.

What ARB Does

• Standardizes publisher-readiness reviews into evidence-backed checks.
• Routes users through a full workflow from first audit to resubmission readiness.
• Supports both fast triage and full-scope audits.
• Produces structured outputs that are easy to reuse in reports, remediation plans, and follow-up reviews.
• Extends beyond AdSense into multi-network readiness, benchmark comparison, and revenue-priority analysis when needed.

Who This Is For

• Site owners preparing for an initial AdSense application.
• Operators trying to reduce rejection risk before resubmission.
• Consultants or agencies auditing multiple publisher sites.
• Teams that need a repeatable audit contract instead of ad hoc reviews.

Why ARB

ARB is designed around three constraints:

1. High-risk policy failures must stop the process early.
2. Audit outputs must be reproducible and evidence-based.
3. Different site types need different emphasis without changing the base scoring contract.

That is why ARB uses:

• 4 review gates.
• 6 scoring pillars: CI, PC, TH, UX, TD, SI.
• 3 score modes: Core 79, Core 79 + Profile, Full 105.
• Veto precedence for critical failures.

Repository Structure

Core Concepts

1. Six Pillars

2. Scoring Scope

3. Veto Rules

Readiness cannot pass if critical blockers fail. At minimum, treat the following as hard-stop conditions:

• TH01: HTTPS not enabled site-wide
• TD01: no privacy policy
• Any unresolved fail in PC items
• Any conflict on affiliate disclosure that cannot be cleared with evidence

See ARB-benchmark.md for the detailed veto contract.

Full Usage Guide

Prerequisites

This repository is documentation-first. It does not ship a standalone CLI or bundled checker runtime. You use it by opening and following the benchmark and skill documents inside your editor or agent environment.

Recommended setup:

• A Markdown-capable editor such as VS Code
• Access to the target site URL, source files, or both
• A way to capture evidence such as screenshots, page links, crawl notes, or extracted findings

Fastest Start

From the repository root:

open ./skills/arb-full-audit/SKILL.md
open ./ARB-benchmark.md
open ./skills/README.md

If you only want a lightweight triage instead of the full workflow:

open ./skills/ads-readiness-assessment/SKILL.md

Standard Workflow

Use this path for most audits:

1. Define the site URL, site type, and desired score mode.
2. Start with skills/arb-full-audit/SKILL.md.
3. Run Gate 1 technical checks.
4. Run Gate 2 content, policy, spam, and copyright checks.
5. Run Gate 3 trust, UX, and disclosure checks.
6. Aggregate scores, apply veto logic, and record top fixes.
7. Use the matching remediation skills.
8. Finish with skills/resubmission-readiness-check/SKILL.md.

Manual Gate-by-Gate Workflow

Use this when you want more control over each phase.

Choose the Right Entry Point

Inputs You Should Prepare

For the best result, gather these before you start:

Supported Review Modes

Most skills can be used in one of these ways:

• URL mode: review the live site.
• Source mode: inspect local project files.
• Manual mode: follow the checklist and record findings yourself.

See skills/README.md for the skill-by-skill mode matrix.

Outputs You Should Expect

A complete audit should normally produce:

• A declared score mode and evaluated scope
• Pillar scores for CI, PC, TH, UX, TD, and SI
• Veto status and any stop conditions
• Evidence-backed findings for non-pass items
• A prioritized fix list
• Suggested next skills
• Optional approval probability and network-gap overlays

Recommended artifact set:

Typical User Paths

Website Owner

1. Start with arb-full-audit.
2. Fix the highest-priority blockers.
3. Run resubmission-readiness-check.
4. Submit only after all veto-adjacent issues are resolved.

Consultant or Agency

1. Use agency-batch-auditor for multi-site intake.
2. Use arb-full-audit for deep reviews on selected sites.
3. Convert findings into client remediation tracks.
4. Re-run targeted checks before delivery or resubmission.

Rejected Publisher

1. Start with rejection-root-cause-analysis.
2. Build the fix sequence with recovery-action-plan.
3. Run the remediation skills that match the failing pillars.
4. Finish with resubmission-readiness-check.

Skill Groups

Orchestration

• arb-full-audit

Diagnostics

• ads-readiness-assessment
• content-audit
• technical-audit
• ux-compliance-audit
• policy-risk-scanner

Remediation

• content-improvement-blueprint
• technical-remediation-guide
• ux-optimization-roadmap
• policy-remediation-plan
• trust-credibility-strategy

Cross-Phase Checks

• affiliate-link-compliance
• copyright-ip-check
• seo-spam-detection

Monitoring and Recovery

• active-compliance-monitor
• health-check-automation
• alert-rules-setup
• rejection-root-cause-analysis
• recovery-action-plan
• appeal-strategy-builder
• resubmission-readiness-check

Scale and Monetization Overlays

• ai-content-compliance
• multi-network-readiness
• revenue-potential-estimator
• geo-localization-compliance
• competitive-benchmark
• agency-batch-auditor

Recommended Reading Order

If you are new to the repository, read in this order:

1. README.md
2. ARB-benchmark.md
3. skills/README.md
4. skills/arb-full-audit/SKILL.md
5. The specific skill docs you plan to run

Working Rules

When using or extending ARB, keep these rules intact:

• Do not invent evidence.
• Do not suppress veto signals.
• Do not change score scope without naming it explicitly.
• Do not mark a site ready just because the weighted score is high.
• Do keep findings reproducible and traceable.

The detailed contributor contract lives in AGENTS.md.

FAQ

Is this a software product or a documentation framework?

It is a documentation-first framework. The repository defines the benchmark, scoring model, skill flow, and output expectations.

Can I use it without evaluating all 105 items?

Yes. Start with Core 79 or Core 79 + Profile when you need a faster or more targeted review.

Should I always start with arb-full-audit?

For most users, yes. Use a specialist skill first only when you already know the problem area.

Does a high score guarantee approval?

No. ARB is a risk-reduction framework, not an official approval predictor.

Related Documents

• ARB-benchmark.md
• skills/README.md
• skills/arb-full-audit/SKILL.md
• readme.zh.md