Evidence-based scan · No signup

ReviewArmor

Audit your Chrome extension before Google does.

AI-coded extensions ship with launch materials that often contradict the manifest — wrong permission claims, fake privacy guarantees, missing disclosures. ReviewArmor scans yourmanifest.jsonagainst known rejection triggers, then generates launch documents grounded in actual evidence — not AI guesses.

Drop your manifest.json here

or

Risk scanning runs locally in your browser. If you generate the Launch Kit, your manifest is sent to our server and AI provider for document generation. We do not store uploaded manifests. Privacy.

Evidence-grounded risk scan

Manifest V3 checks covering broad host access (incl. content_scripts.matches), remote code, sensitive APIs, OAuth scopes, permission redundancy, page overrides, and missing assets.

5 contradiction-checked documents

Privacy policy, permission justifications, store listing, reviewer notes, dashboard fields — each verified against your manifest to catch AI hallucinations before submission.

Auto-retry on factual errors

If the AI writes 'does not request all_urls' but your manifest does, we catch it and regenerate. No more rejection from your own materials.

Why this, when Google ships a CHROMEWEBSTORE.md generator?

Google's 2026-05-19 AI coding skill generates dashboard files from templates. ReviewArmor does the opposite — it auditswhat you (or your AI agent) already wrote, finds places where the generated docs contradict your manifest, and rewrites them with concrete evidence. Reviewers reject extensions whose listings claim things the code can't back up.