Better Writing by @jakubkrehel · By Sergey
Get Page Writing Review > 60 seconds
Comprehensive copyediting based on academic and stylistic principles to enhance clarity, conciseness, and impact.
Runs free on Gemini 3 Flash Preview — sign up to try any other model.
Run 4 times so far
From a real result — run it to see the rest.
Executive Summary
The landing page for PromptHub features strong, benefit-driven messaging and a clear value proposition. However, the copy is inconsistent in its capitalization policy, occasionally uses passive or "we" phrasing in ways that distance the reader, and includes several placeholder-style strings and generic links that do not clearly describe their destination. The tone is generally appropriate (Professional/Neutral), but structural polish is needed to ensure the interface "disappears" and remains invisible to the user.
Findings by Principle
1. One Capitalization Policy
The site oscillates between Title Case and Sentence Case for headings and buttons (e.g., "Level up your prompt management" vs. "Run Prompt"). Sentence case is more scalable for localization and provides a calmer reading experience.
Navigation & Hero Buttons
- Current Text: "Start for free" | "Schedule a demo"
- Proposed Text: "Start for free" | "Schedule a demo" (Note: Keep as is, but ensure all other buttons match).
- Severity: LOW
- Rationale: Establishing a baseline. Some site elements use sentence case correctly while others do not.
API Tabs
- Current Text: "Run Prompt", "Return Prompt", "Return All Prompts"
- Proposed Text: "Run prompt", "Return prompt", "Return all prompts"
- Severity: LOW
- Rationale: Violates the capitalization policy. Mixed casing across buttons ("Start for free" vs "Run Prompt") reads as sloppiness.
2. Say "You", Never "The User"
The copy occasionally slips into "we" phrasing or speaks about the platform in a way that creates friction.
Documentation Links
- Current Text: "See how you can automatically evaluate prompts"
- Proposed Text: "Evaluate prompts automatically"
- Severity: LOW
- Rationale: Focus on the action rather than the meta-description of what "you can" do.
API Description
- Current Text: "Retrieve prompts from any branch, inject variables with your app data, and pass along any metadata."
- Proposed Text: "Retrieve prompts from any branch, inject variables with your data, and include metadata."
- Severity: LOW
- Rationale: "Your app data" is redundant; "your data" or just "data" is leaner.
3. Plain Words Over Clever Ones
Several areas contain "placeholder" text that was left in the production environment.
Prompt Library (Trending)
- Current Text: "Prompt Generator. 7 prompt framework description placeholder"
- Proposed Text: "A flexible framework for generating complex multi-step prompts." (Or representative description)
- Severity: HIGH
- Rationale: Displaying "placeholder" to a user signals an unfinished product and breaks trust.
4. Links Describe Their Destination
Generic "Learn more" style links or links with repetitive lead-ins make navigation difficult for screen readers.
Feature Links
- Current Text: "See how versioning works", "AI tools for prompt engineering", "How to chain prompts"
- Proposed Text: "Read about prompt versioning", "Explore AI prompt tools", "Read the prompt chaining guide"
- Severity: MEDIUM
- Rationale: Links should make sense out of context. "How to chain prompts" is good, but "See how versioning works" is slightly passive. Suffixing them ensures uniqueness.
5. Verb-First Buttons
Most buttons are well-actioned, but the ProductHunt banner is a social invitation rather than a clear action.
ProductHunt Banner
- Current Text: "We're live on ProductHunt, come say hi!"
- Proposed Text: "View PromptHub on ProductHunt"
- Severity: LOW
- Rationale: "Come say hi" is slightly colloquial/clever; name the specific destination.
Verification Checklist
- Core Principles Check: Verified against all 11 principles including "You" over "User" and Verb-first buttons.
- Capitalization: Checked headings against buttons; found significant inconsistency.
- Variable Interpolation: Checked "Prompt Generator" module; found a "placeholder" string error.
- Device Verbs: Confirmed no pointer-specific ("click") language is used in the main body.
- Empty States: Not applicable (Landing page content present).
- Error Messages: Not applicable (No form errors triggered in the source text).
Verdict
VERDICT: Needs changes
Reasoning: While the marketing copy is effective, the presence of the word "placeholder" in the Trending Prompts section is a High Severity finding that must be addressed. Additionally, the inconsistent capitalization across the API documentation tabs and primary buttons needs to be unified to meet a professional standard.
Needs changes
How it works
From blank fields to a finished result
Fill in one field.
Site page — that's everything Better Writing by @jakubkrehel needs.
AI generates it instantly.
Better Writing by @jakubkrehel runs the prompt behind the scenes and returns a clean, formatted write-up in seconds.
Copy it, or share the link.
Every run gets a shareable link — anyone can view the result, no login required.
FAQ
Frequently asked questions
Is this free to use?
Yes — running Better Writing by @jakubkrehel is free, and you don't need an account.
Do I need to sign up?
No. Fill in the fields above and run it. Signing up only matters if you want to build or save prompts of your own.
How is the result generated?
Better Writing by @jakubkrehel runs on FirePrompts — your inputs fill in a proven prompt template, which an AI model then completes.
Can I share my result?
Every run gets a shareable link, so anyone can view it — no login required.
Build tools like this for your team
Version every prompt, run any model, and ship your own public tools. Free during early access.