Content Lab (RAG Workbench)

How to use the Split-View Workbench to analyze, rewrite, and score your content for AI citation readiness — and how to interpret the Before/After Score Delta.

The Content Lab works on any publicly accessible URL. You do not need an active project selected — it is a standalone workbench. Paste any URL to begin.

Overview

The RAG Optimization Workbench is your content transformation studio. It fetches any page, scores it against the same 4-pillar AI Readiness model used in the Audit, and then offers a one-click AI rewrite that is specifically engineered to improve RAG (Retrieval-Augmented Generation) retrieval performance.

The key difference from the standard AI Readiness Audit: the Content Lab is optimized for before-and-after comparison. You can see exactly how much each pillar score improves after an AI rewrite — before committing the changes to your live page.

Credit for Deep AI Analysis

Credits for RAG Rewrite

Pillar scores shown side-by-side

Split-View Workbench

The workbench is divided into two panels that work together:

Left panel — Input & Analysis

Enter your URL here and run the Deep AI Analysis. The Before scores appear in this panel after the analysis completes. This panel also contains the "Rewrite for AI" button that triggers the AI transformation.

Right panel — Before / After Scores

Once an analysis has run, this panel displays your Before scores. After a rewrite, the After scores appear alongside the Before scores for direct comparison. Each pillar shows the raw score change and the percentage improvement in citation probability.

Run Deep AI Analysis

The Run Deep AI Analysis button costs 1 credit per use. It fetches your page and scores it across all 4 pillars, establishing the Before baseline.

Paste your URL

Enter a full public URL (including https://) in the input field at the top of the workbench. The page must be accessible without a login.

Click Run Deep AI Analysis

The analysis begins immediately. The button shows a loading state while the page is fetched, parsed, and scored.

Review your Before scores

Once complete, your AI Readiness Score and all 4 pillar scores appear in the right panel as your Before baseline. These scores are the benchmark you are trying to improve.

Decide whether to rewrite

If your overall score is below 70 or any pillar score is weak, click Rewrite for AI to generate an optimized version of your content.

Run the analysis before the rewrite

Always run Deep AI Analysis first. The Before scores are required to compute the After delta — you cannot see the improvement comparison without an established baseline.

Before/After Score Delta

After you run a rewrite, the workbench automatically re-analyzes the rewritten content and computes an improvement delta for every pillar. Each pillar displays a Delta Badge showing:

Badge format	What it means
+12 pts · +18% Citation Probability	Pillar score improved by 12 points; citation probability for that pillar increased by 18%
+0 pts · No change	Pillar score was already optimal; no improvement possible for this dimension

The Overall Score delta is shown prominently at the top — this is the single most important number. It tells you the total citation readiness improvement the AI rewrite achieved.

Delta is computed from the rewritten content, not a prediction

After a rewrite, the workbench runs a full re-analysis pass on the transformed text. The delta is not an estimate — it is the real score difference between the original and rewritten content.

Rewrite for AI (1 credit)

The Rewrite for AI button costs 1 credit per use. It sends your page content to Gemini with a structured prompt engineered specifically for RAG retrieval performance — not general readability or SEO.

The rewrite optimizes for:

Clearer H2 headings phrased as questions that match AI query patterns
Concise, definition-first paragraph structure that AI engines can extract cleanly
Numbered lists and step-by-step flows for instructional content
Metadata signals (title and description) aligned to the page's primary topic intent
Citation anchor patterns (explicit external references, author attribution cues)

Run Deep AI Analysis first

Ensure a Before score baseline exists. The Rewrite for AI button is enabled only after an analysis has completed.

Click Rewrite for AI

The button deducts 1 credit and sends your page content to the Gemini rewrite engine. A loading state is shown while the transformation processes (typically 10–20 seconds).

Review the rewritten content

The transformed content appears in the workbench. Scroll through it to review the structural and language changes the AI made.

Check the After scores and delta

The right panel updates automatically with your After scores and the Delta Badge for each pillar. Compare each pillar to decide which rewrites to prioritize implementing.

Copy and implement the rewritten content

Copy the rewritten sections that showed the highest delta improvement. Implement them on your live page. Re-run the AI Readiness Audit on the updated URL to confirm the score improvement.

Content Lab actions require credits

Both Run Deep AI Analysis (1 credit) and Rewrite for AI (1 credit) consume credits. Free tier users must have sufficient credit balance to use either action. With a BYOK Gemini key added in Settings → API Keys, both actions cost 0 credits.

The 4 scoring pillars

The Content Lab uses the same 4-pillar model as the AI Readiness Audit, displayed as individual score cards:

Pillar	What it scores	Rewrite impact
Structure	H1/H2 hierarchy, lists, question-phrased headings	High — rewrites restructure headings and add list patterns
Metadata	Title tag, meta description, Open Graph alignment	Medium — rewrites suggest optimized title and description
Content Quality	Readability, sentence length, passive voice ratio, word count	High — rewrites reduce passive voice and improve clarity
Citation Patterns	Schema markup, external links, author attribution cues	Medium — rewrites add citation anchor language; Schema requires separate fix

Use the Content Lab alongside the AI Readiness Audit

The Content Lab improves your content. The AI Readiness Audit generates the exact Schema JSON-LD code to fix your Citation Patterns pillar. Use both together: Content Lab for the writing, Audit for the Schema.

Recent Sessions history

The workbench stores your last 5 analysis sessions automatically. The Recent Sessions panel shows:

The domain and page analyzed
The Before score badge for that session
The After score badge (if a rewrite was run)
The delta improvement between Before and After
The date the session was run

Click any session row to reload that analysis into the workbench. This lets you revisit the rewritten content and delta scores without re-running the analysis.

Sessions are stored per account, not per project

Recent Sessions are not scoped to a specific project. They show the last 5 analyses across all URLs you have processed in the Content Lab, regardless of which project is active.

Cost & credits

Action	Credit cost	BYOK bypass
Run Deep AI Analysis	1 credit	Yes — add Gemini API key in Settings → API Keys
Rewrite for AI	1 credit	Yes — add Gemini API key in Settings → API Keys

To run unlimited AI rewrites at zero credit cost, add your Gemini API key in Settings → API Keys. See the BYOK guide for step-by-step instructions.

Was this page helpful?