You're a founder, blogger or growth marketer shipping AI-assisted content. You don't need more text โ you need to know, before you publish, whether a draft clears the quality bar, and exactly what to fix first if it doesn't. Every piece gets a Content Quality Score on four transparent dimensions, a diagnosis of what held it back, the single highest-priority fix, and three independent QA checks (AI-Tone phrasing, evidence needs, publishing readiness). Paste any article into the Analyzer to run the same gate on a competitor's draft โ read how to check AI content quality before publishing or what 29 scored drafts revealed.
One score. Four visible reasons. No black-box "quality" number โ and the same rubric runs in generation and in the Analyzer.
Non-obvious units a reader couldn't guess from the title โ the part that wouldn't exist if the draft had been written from the headline alone.
Numbers, names, dates, conditions and steps in place of vague qualifiers โ so a reader can act on the piece instead of nodding at it.
Whether the claims that need attribution carry it, instead of asserting figures with nothing behind them.
Mechanism and reasoning instead of description โ the difference between "do this" and "do this because of that".
Each dimension keeps its own deduction note, so you can see why a score landed where it did. The total is compared against the pass line for your industry โ strong-template industries hold an 8.0 bar, mid industries 7.5โ7.8, weaker ones 7.0 โ and drafts below the bar are rewritten before delivery rather than shipped quietly.
Scoring is built to be boring and trustworthy, because that's what a quality gate has to be:
See it on a real draft: a real before/after report โ 6.2 → 7.6 โ
Three independent checks run alongside the score. None of them changes your Content Quality Score โ that number only moves when the content moves.
A lint pass flags phrasing patterns โ "you'll" chains, roadmap language, scaffold leakage โ with the exact phrases listed. It reports phrasing only: it never rewrites your facts, your claims or your score.
Claims are classified by what they need (needed / recommended / not needed) and pointed at by sentence. It flags evidence needs โ it does not certify that a claim is true, and we don't claim it does.
Pick the platforms you plan to publish on and the draft is checked against each one's structural expectations, so you find the gap before the platform does.
AI-Tone is a separate signal. Cleaning mechanical phrasing does not raise or lower the Content Quality Score โ the score reflects the substance of the draft, not its accent.
This is the part most tools skip: turning a number into a publishing decision.
Four dimensions, median of three samples, against your industry's pass line.
Every dimension keeps its own deduction note โ nothing hidden behind one number.
One issue, the exact sentence it points at, one recommended action โ not a checklist of thirty.
AI-Tone ยท Evidence ยท Publishing readiness โ parallel reads that never fold into the quality score.
Edit, then score again: before/after on the same rubric, so improvement is measured, not assumed.
The gate gives you the read; the call stays yours.
Top priority: the four-item checklist is templated content inside a format skeleton โ every item opens with the same imperative construction, so the section reads as generated rather than reasoned.
The sentence it points at"Read your first sentence. If it could open any article on this topic, it carries no information. Name a number, a name, a date, or a decision."
A competitor's post, a viral piece, or your own draft. You can judge content before you create more of it.
One credit analyzes any article from any site or creator. Every form gets a Content Quality Score: the 4 dimensions for long-form, the short-form rubric for social posts, and script quality (hook / retention / CTA / clarity) for video scripts. Same rubric, same gate โ generation and Analyzer never use two different rulers.
AI-tone flags are checked immediately at zero cost before the paid analysis โ you see phrasing red flags in seconds and only spend a credit when the piece is worth scoring.
Every analysis returns the single most actionable problem โ quoted verbatim from your text, with a concrete instruction โ so "Fix this article" turns analysis into a rewrite plan, not just a number.
Score a piece that's working, then "Generate similar article" carries the topic, platform and the winning patterns into the generator.
The quality standard doesn't change. What changes is the format, the structure and the publishing context.
Real output you can open right now: a live Google long-form article ยท a real before/after X report (6.2 → 7.6) ยท 35 QC-scored topic drafts
Generation is the capability, not the pitch: the same gate scores what comes out.
8 platforms and 24 native formats. The prompt architecture switches to match that platform's publishing context.
One credit produces a full article, script or post. If the draft misses your pass line, it regenerates free until it clears โ you don't spend a second credit.
Every article ships with the 4-dimension diagnostic, the ForgeRank pass line it cleared, the separate QA checks, and the Deep Insight Report explaining the structural choices.
What the generator starts from โ so the gate isn't the only thing standing between you and a generic draft.
Templates that already scored diag โฅ 8.5 (the 8.5 collection bar, above the 8.0 production pass line) are available as starting recipes rather than as a black box.
H2 blueprints drawn from top-performing pieces in that industry are injected into the recipe, so drafts follow a structure that has already earned the read โ not generic advice.
Everything you paste into the platform, plus the parts creators forget.
Alternate titles, caption/title variants and hashtag sets for the platform you're posting to.
A posting-time recommendation with the reasoning attached, so you can sanity-check it instead of trusting a number.
Reply starters and a pinned comment, plus sourced pull quotes taken from your article โ so you never copy-paste an unverified claim.
Not sure what to write about? The Trends page surfaces rising topics and demand signals across niches — updated every 6 hours from live sources, with real counts where available and honest fallbacks clearly marked. Pick a trend, generate content for it, then run it through the same gate.
See today's trends โNo card required. Paste a draft and see the score, the diagnosis and the one fix that matters most โ or pick a platform and generate.