The Compliance Layer for CPG Brands
EVERY FINDING, WITH ITS CITATIONCLEAR TRACEABLE YOURS TO RUN
Paste your label and get every federal, state, retailer, claims and ingredient finding, each with the citation behind it, then the corrected copy to hand your designer.
→
Try The Engine On Sample Labels
Walk a sample label through the engine. No account needed: watch every violation surface; create a free account to see them fixed.
→
What Is CPGShield?
A DIY label-compliance engine: what we check, what you get, and what it costs. Scanning is free.
Sample output: a real scan returns every finding with its citation, and the corrected label copy.
CPGSHIELD
SIGN IN
Access your compliance workspace
or
New here? Register with your email address.
✉
CHECK YOUR EMAIL
We sent a verification link to
Click the link to activate your account. Once verified, you can sign in. If this address already has a CPGShield account, no new one was created. Sign in with it instead, or reset the password if you have forgotten it.
TWO-FACTOR AUTHENTICATION
Enter the 6-digit code from your authenticator app.
WHAT IS CPGSHIELD?
The label compliance engine CPG brands run themselves, no consultant required.
The short version▾
Paste your label PDF or text. CPGShield checks it against every rule we encode: federal FDA labeling, MoCRA, state ingredient bans, Prop 65, PFAS restrictions, FTC claims standards, and the published clean standards of the retailers you sell to.
You get back every requirement we screen that your label violates, each one named with the regulation behind it and the severity it carries. Then you fix them with a few clicks: compliant wording, required warnings, and corrected label copy ready to hand straight to your designer.
You run it yourself. No consultant to engage, no back and forth, and no demo call before you can find out whether your label has a problem.
Why this exists▾
Label compliance is not one problem you solve once. It is a problem that returns with every SKU, every reformulation, and every new state law that takes effect.
The rules are scattered across federal regulation, a growing patchwork of state chemical laws, and retailer standards that are published policy rather than law and change without notice. Nobody hands you a single list. A brand owner writing a label is expected to know all of it, and most of them find out what they got wrong from a retailer’s compliance team after the print run.
The existing options are a regulatory consultant at a few hundred dollars per label, or formulation software built for cosmetic chemists at enterprise prices with a sales call attached.
Neither of those helps you at 10:30 on a Tuesday when the label goes to print Thursday.
How it worksyou only pay when there is something to fix▾
1. Scan, free. In our label scanner, paste the PDF or text from your label. The engine runs every rule we encode and returns a verdict plus the exact count of violations and warnings. Free, every time, no card, no limit.
2. See the verdict. If your label is clean by CPGShield standards, you are done. You pay nothing.
3. Unlock to see the details. If something is wrong, one label credit unlocks that label permanently: every finding named and explained, the fixes, the rejection diagnosis, and the export.
You only pay when there is something to fix.
What you get when you unlock a label▾
Every finding, with the law behind it. Each violation names its source: a CFR section, a state statute, or a retailer’s published standard. Not a vague warning. The actual citation, quoted and explained in the app, so you can see exactly what the requirement says without hunting through a government website.
Severity that means something. Findings are separated into what blocks you and what is advisory, grouped into six categories with one clear compliance gate. You always know which items stop a print run and which are judgment calls.
A record of what ran. Every scan lists every check it performed: what fired, what passed, what did not apply to your product, and what it could not check from label text alone. When something falls outside what a text scan can determine, we say so instead of reporting it as a pass. You get a result you can reason about rather than a black box verdict.
Corrected copy, not merely a diagnosis. Most compliance tools tell you what is wrong. CPGShield hands you the fix. Illegal claims get compliant replacement wording. Missing warnings, net quantity statements, the manufacturer line, and the Prop 65 warning are all supplied in the correct regulatory format, ready to paste into your artwork.
The tools▾
Fix your label
Apply All Fixes. One guided pass rebuilds your label text with every fixable problem corrected. Illegal claims get compliant replacement wording. Banned ingredients come pre-checked for removal, while conditionally restricted ones are left to your judgment. Missing warnings, net quantity, the manufacturer line, and the Prop 65 warning are all handled. Run it, review, run it again.
Label Builder. Start a compliant label from scratch instead of retrofitting one. Useful when the formula exists but the packaging copy does not.
Handle a retailer
Retailer Rejection. When a buyer or a retailer’s compliance team rejects your label, paste their feedback: CPGShield diagnoses exactly what they flagged, maps each objection to the requirement behind it, and drafts your appeal letter. Included with every label unlock at no extra charge.
Compliance Matrix. A requirement by requirement view across states and retailer programs, so you can see where a single SKU stands everywhere at once.
PDF compliance report. Export any scan as a report you can hand to a buyer, a broker, or a retailer’s compliance team.
Check before you commit
Ingredient Check. Paste your ingredient list and see every restriction we hold for each ingredient, grouped by jurisdiction. Useful before you commit to a formula rather than after.
Marketing Claims Check. Test a claim before it goes on the label or into an ad. It screens for drug claims, unsupported language, and phrasing that regulators and retailers reject.
Claims screening. Your marketing copy is screened alongside the label panels automatically on every scan, because the line between a cosmetic claim and an unapproved drug claim is one many brands cross without noticing.
Stay organized
Tasks. Every violation can become a tracked remediation task with status and assignment. Free on every account.
The research tools (Ingredient Check, Marketing Claims Check, Label Builder and Compliance Matrix) open with any credit pack. Scanning stays free.
What we cover▾
- 21 CFR cosmetic labeling
- MoCRA
- State chemical and ingredient laws
- Prop 65
- PFAS restrictions
- FTC claims standards and Green Guides
- Retailer clean beauty programs
Every count we publish is computed live from the rule base. No coverage number on this site is typed by hand.
USA COSMETICS · SCREENING TOOL · NOT LEGAL ADVICE
SEE IT WORK
Five sample labels, from a near-miss to a stress test, scanned by the real engine
Loading samples...
SAMPLE LABELS ARE SYNTHETIC · RESULTS COME FROM THE REAL ENGINE
Dashboard
Welcome back
-
Labels
-
Total Scans
-
Passing Scans
-
Failed Scans
-
Health Score
-
Compliance Tasks
RULE COVERAGE
Coverage & Methodology →
Loading...
PROJECTS
Loading...
Projects help you organize multiple SKUs. You can create one after your first scan.
RECENT SCANS
Loading...
Labels
Every label you've scanned, status, full scan history, and workflow, in one place
Loading labels...
Label
Loading...
Loading compliance items...
New scan
Paste your label or upload a photo for a full compliance check. Scanning is always free.
Include: product name · net quantity · full ingredient list (the ingredient names exactly as they appear on your label) · manufacturer name & address · any warnings. The more complete the text, the more checks we can run.
Click to upload or drag & drop · JPEG · PNG · TIFF
Upload your print-ready label PDF · we measure net-quantity, warning-statement, ingredient, and identity formatting against 21 CFR 701.13, 701.3, 701.11 & 740.2
Vector (print-ready) PDFs only, a photo saved as PDF can't be measured. We read the first page as the front panel at actual print size.
Formula concentrations (optional)
Label builder
Build a correctly structured cosmetic label step by step
PRODUCT IDENTITY
STEP 1 OF 5 · Required under 21 CFR 701.11
NET QUANTITY
STEP 2 OF 5 · Required under 21 CFR 701.13
INGREDIENTS
STEP 3 OF 5 · Required under 21 CFR 701.3
List in descending order of predominance using INCI names, separated by commas. Ingredients at ≤1% may be listed in any order at the end.
MANUFACTURER INFO
STEP 4 OF 5 · Required under 21 CFR 701.12
WARNINGS & DIRECTIONS
STEP 5 OF 5 · Best practice + category requirements
FAIL
SCAN UNSUPPORTED
Rule library
Loading...
Loading rule library...
Loading ingredient hazards...
Rejection mining (internal)
PRICING
Catch every issue we check for — before the print run, and before a retailer does. Pay once per label, never per month.
Paste a label draft and see every violation we find, with the fixes.
Catch the problems while they still cost nothing to fix, instead of after a print run or a retailer rejection.
CPGShield doesn't replace your regulatory consultant. It means you walk into your regulatory review with a clean label,
not a pile of basic errors.
Start free
Unlimited scans, always. Every scan shows you what's wrong: issue counts, severities, and category breakdown. Unlock a label to see every fix, the corrected-label guide, and PDF export.
Credit packs
1 label credit unlocks 1 label: fully, forever. Re-scan as you revise it at no extra charge. Every pack also includes check credits: one credit runs the Ingredient Check, Marketing Claims Check, or Compliance Matrix once. Credits never expire, and your price is locked in at purchase. Bigger packs cost less per label and carry more checks per label.
Buy credits
Loading pricing…
Unlimited free scans · credits never expire · price locked at purchase · no subscriptions, no contracts. Payments processed securely by Stripe.
FREQUENTLY ASKED QUESTIONS
You can scan unlimited labels for free. Every scan shows issue counts, severities, and what's wrong. To see the fixes, the corrected-label guide, and PDF export for a label, unlock it with a label credit (1 label credit = 1 label, unlocked forever). No setup call required.
One label credit covers one product (one SKU), fully unlocked forever, including every re-scan as you revise that label at no extra charge. Materially different products require separate labels and separate credits.
No, and we'd never claim to be, but for most of what a regulatory review catches, it does the same first-pass work automatically. A consultant reads your label line by line against the CFR, the state ingredient bans, the retailer lists, and the claims rules, and comes back with a marked-up list. That first read is exactly what CPGShield automates: it checks the same rule set, cites the same regulations, and hands you corrected copy for everything it can fix. What it does not do is exercise judgment on the genuinely ambiguous calls, the borderline claim, the novel ingredient, the strategic question of how aggressively to position a product. That's where a human belongs. Use CPGShield so you walk into that review with a clean label instead of a pile of basic errors, and so your consultant's hours, and your budget, go to the decisions that actually need a person, not to catching a missing net-quantity statement or a paraben on a retailer's banned list.
Federal labeling under FDA cosmetic regulations and the Fair Packaging and Labeling Act; state cosmetic ingredient restrictions across every state we cover, including Prop 65, PFAS bans, and the phthalate and formaldehyde-releaser laws that are reshaping the market; retailer clean-standard programs for Amazon, Sephora, Target, Walmart, Ulta and Whole Foods Market, screened against their published banned-ingredient lists; marketing-claims risk, drug claims, pesticide and FIFRA claims, FTC "Made in USA" and Green Guides environmental claims, cruelty-free and USDA-organic claims, and dozens of misleading-language patterns; and ingredient hazard flags spanning heavy metals, endocrine disruptors, restricted preservatives, and fragrance allergens. Every single one is tied to the CFR section or statute behind it, nothing is a black box, and you can open the complete, always-current list any time in our Rules Library. We deliberately do not dump the entire Code of Federal Regulations on you, the CFR runs thousands of pages, most of it irrelevant to a cosmetic label. We encode only the rules that actually apply to a label and that a text scan can meaningfully check, so a flag from CPGShield is a real, relevant problem, not regulatory noise. EU requirements are available as an opt-in mode.
Three things you can check yourself, rather than a number you have to take on faith. First, traceability: every rule points to the primary regulation behind it, the actual CFR section or statute, not a paraphrase, so you, or your attorney, can check any finding against the law yourself. We re-verify the rules the engine leans on most against those primary sources directly, and the high-consequence ones, the outright federal ingredient prohibitions and the in-force state bans, we have confirmed line by line against the government's own published text. Second, we tell you what we do not check: every scan carries a plain list of what a label-text screen cannot determine, including formula safety, lab testing, trace contaminants and facility registration. A tool that only tells you what it found gives you no way to judge what it missed. Third, we never report a skipped check as a pass: every check is recorded on its own, as exactly one of four things, it found a problem, it found no problem, it did not run and why, or it broke while running. A check with no record at all is treated as not run. When we cannot read a usable ingredient list, every check that searches it is reported as not run, with the reason. Absence of a finding is never presented as evidence of absence.
Two principles run the whole library: relevance and traceability. On relevance, we curate hard. The universe of cosmetics regulation is enormous and most of it never touches a retail label; we encode only the rules that apply to a label and that a text scan can meaningfully evaluate, so what you see is signal, not noise. On traceability, every rule carries the CFR section or statute it comes from, laid out in the open in the Rules Library, because a compliance tool you can't audit is one you shouldn't trust. Currency is built into the engine itself: when a state ban has a future effective date, we flag it as a warning until that date and flip it to blocking the day it takes force, so you see the rule as of the last-updated date shown on the library, and we verified the effective date and severity of every in-force state ban against the statute directly. We re-check the rules the engine relies on most against primary sources. And we're honest about maturity: our ingredient-hazard entries are seeded from authoritative sources and carry a visible verification status in the Rules Library, so you always know exactly which rules have cleared which stage of review rather than being asked to assume they all have.
Anything that can't be determined from label text alone: formula safety substantiation, ingredient concentration limits, trace contaminants, stability and microbiological testing, facility registration status, Good Manufacturing Practice compliance, physical label dimensions and font size, and whether your claims are factually true for your product. Where these matter, we flag them for your review rather than guessing. Non-US regulations are out of scope except the EU opt-in mode.
Credits never expire, and the price you pay is locked in at purchase. Buying a credit pack does not create a subscription, a recurring charge, or an auto-renewal: you buy exactly the labels you need. A small number of accounts opened before July 2026 hold a legacy Pro subscription from a plan we no longer sell; those continue to renew on their existing schedule until cancelled, self-serve, from Settings.
Your label text and scan results are stored securely, never sold, and never shared for marketing. We use trusted service providers (like Stripe for payments) to operate the product. See our Privacy Policy for full details.
Unspent credits are refundable within 7 days of purchase. Once a credit is spent (a label unlock or a check run), that use is permanent and the credit is consumed. Contact support@cpgshield.com.
1
2
3
Tell us about your brand
Helps us tailor compliance checks to your products.
Your brand
Your site makes claims too. We check that what you say online matches what your label can legally support.
What you sell
Say it in your own words. If it is something we do not cover yet, we would rather tell you that up front than scan it badly.
Drives the MoCRA responsible-person check.
Where you sell
Amazon, Shopify, TikTok Shop or anywhere else that ships to all 50 states. If this is on, every state's rules apply to you.
Tells us whether EU cosmetic rules are worth showing you. They stay off for US-only scans.
Your business
Run your first scan
Choose how you'd like to get started. You can always run more scans later.
PASTE TEXT
Copy your label text directly from a document or packaging spec
UPLOAD IMAGE
Upload a label photo or product image, we extract text via OCR
USE SAMPLE
Try CPGShield with our pre-loaded "Truly Brightening Body Serum" label
Click to upload · JPEG · PNG · TIFF
-
findings found on your first scan.
Settings
Account and workspace preferences
ABOUT CPGSHIELD
CPGShield exists to make regulatory clarity something a growing brand can actually reach. Cosmetic labeling in the United States is governed by an expanding patchwork, federal FDA requirements layered over a widening field of state ingredient laws, layered again over retailer standards that move faster than either. Most founders meet that patchwork for the first time when something has already gone wrong: a delisted product, a rejected submission, a print run that has to be scrapped. We built CPGShield to move that moment earlier, back to the point where a change is still cheap and still yours to make.
Our approach is to encode the rules that genuinely govern a label, and then to show our work. Every check is tied to a specific CFR section, statute, or published retailer standard, and every finding names the authority behind it, so nothing arrives as a verdict you cannot trace. We aim to match the shape of the law rather than the shape of a sales sheet: where a legislature writes a class-level ban, we screen the whole class, and where regulators enumerate substances, we screen them by name. That is why, alongside our total rule count, the table on our Coverage & Methodology page reports what each statute reaches: a single class rule can cover an entire chemical family, and counting it as "one" would tell you the opposite of the truth.
What we are building toward is a working layer between a brand's label and the rules that apply to it, maintained continuously as those rules change, because they do, constantly. That page is part of that intent: coverage stated plainly, sources cited openly, and the edges of our scope drawn where they actually fall, with a route forward for each one. CPGShield is a screening and drafting tool designed to support your compliance process and the professionals who advise you. It is not legal advice, and it is not a substitute for qualified review.
SAMPLE DATA
Explore with sample labels
Adds 3 pre-scanned sample labels to your account, one passing, one with minor issues, one failing, so you can explore the full workflow before committing real label data. Disabling removes all sample labels and their scans.
Disabled
BRAND PROFILE
Brand name & category
No brand profile set.
LEGAL & METHODOLOGY
Terms of Service
The agreement covering your use of CPGShield, screening-tool scope, credits and refunds, and your responsibilities.
Coverage & Methodology
What CPGShield checks, how rules are sourced and verified, and what stays out of scope.
Privacy Policy
What we collect, how we use it, the third parties we share it with, and your access, deletion, and portability rights.
Accessibility Statement
Our WCAG 2.1 AA target, what's in place today, and how to report a barrier.
ACCOUNT
Signed in as
-
Two-factor authentication
Checking status…
Invite a teammate
Seat management is coming in a future release.
SHARE FEEDBACK
SIGN OUT
Sign out of this account
Ends your session everywhere — this device and any other you're signed into. Nothing is deleted: your labels, scans, unlocked results and credits are all waiting when you sign back in with the same email.