GitHub Copilot Prompt for Text Correction
GitHub Copilot, GitHub's AI assistant integrated directly into your code editor, is not limited to code generation. It also excels at correcting and improving texts, whether technical documentation, README files, code comments, or editorial content. By leveraging its linguistic understanding capabilities, you can entrust it with proofreading your texts to identify and fix spelling, grammar, syntax, and style errors. This approach is particularly valuable for developers writing multilingual documentation or clear commit messages. Instead of switching to an external tool, you stay in your usual work environment, streamlining your workflow. The prompt we offer here is structured to guide Copilot precisely: it defines the expected role, the types of corrections desired, and the output format, ensuring consistent and actionable results every time.
Paste in your AI
Paste this prompt in ChatGPT, Claude or Gemini and customize the variables in brackets.
Act as a professional Francophone proofreader. Analyze the following text and correct all spelling, grammar, conjugation, punctuation, and syntax errors. Preserve the original meaning, tone, and style of the author. For each correction made, add a brief comment in square brackets explaining the nature of the error. First return the corrected text in full, then a summary list of changes made. Text to correct:
[PASTE_YOUR_TEXT_HERE]
Personalize this prompt with Léa
Answer 3 questions and Léa tailors the prompt to your situation.
Why this prompt works
This prompt works by assigning an explicit expert role (professional proofreader), which directs the model towards a rigorous linguistic register. The exhaustive list of error types to target avoids omissions, while the request for comments in square brackets forces a reflective analysis that improves correction accuracy. The dual output format (corrected text + summary) allows for quick verification without re-reading the entire document.
Use Cases
Variants
Expected Output
You will get a clean, corrected version of your text, faithful to the original style and tone, accompanied by inline annotations indicating each modification. A summary at the end of the response will list all corrections with their justification, allowing you to learn from recurring errors and validate each change at a glance.
Frequently Asked Questions
Can GitHub Copilot correct texts in languages other than French?
Yes, GitHub Copilot effectively handles corrections in most European languages as well as Chinese, Japanese, and Korean. You just need to adapt the prompt by specifying the target language. However, its performance is best in English and French, the languages its model was most trained on. For less widely covered languages, it's recommended to verify corrections with a native speaker.
Does Copilot respect the style and tone of my original text when making corrections?
When the prompt explicitly specifies it — as in our main version with the instruction 'preserve the author's tone and style' — Copilot strives to maintain the text's original voice. Without this instruction, it may tend to rephrase some passages in a more neutral style. For texts with a very strong tone (humorous, literary, technical), specify the expected register in your prompt for best results.
What is the maximum text length I can have corrected in one go?
The limit depends on the context window of the model used by Copilot. In practice, you can process texts of 2,000 to 4,000 words in a single pass via Copilot Chat. For longer documents, it's better to split the text into logical sections (chapters, thematic paragraphs) and submit them separately. This ensures better correction quality because the model can fully focus on each segment.
Learn more
Check the full skill on Prompt Guide to master this technique from A to Z.
View on Prompt Guide📬 Get new prompts every week
Join our newsletter and never miss a prompt.
Similar Prompts
Generate Batch SEO Meta Descriptions
Generates unique SEO meta descriptions and title tags in batch for multiple pages, with character count and short variant.
Perplexity Prompt for Writing a Press Release
Writing an effective press release requires mastery of journalistic codes, a rigorous structure, and a compelling hook. Perplexity, thanks to its real-time research and intelligent synthesis capabilities, becomes a valuable ally for this task. Unlike a simple text generator, Perplexity can cross-reference current sources, verify facts, and produce a press release grounded in current media reality. Whether you are announcing a product launch, a funding round, a strategic partnership, or an event, the tool helps you structure your message according to PR standards: inverted pyramid, executive quote, data, and clear call to action. By crafting a precise prompt, you get a professional first draft that you can refine in minutes, whereas traditional writing would require several hours of research and formatting. This page offers optimized prompts to get the most out of Perplexity for writing your press releases, regardless of your level of communication expertise.
Write Your Professional Biography
Creates 4 professional biography versions adapted to Twitter, LinkedIn, conferences, and web pages, with optimal usage context for each.
B2B Prospecting Cold Email That Gets Replies
Generates 3 cold email B2B variants with different approaches (problem, value, curiosity), 3 testable subject lines, and Day+5 follow-up for each.