Developer builds ruleset to strip AI writing patterns from AI-generated text

·1 views

A developer published a post on DEV Community describing a tool called Alchemy, a Markdown-based ruleset designed to remove telltale AI writing patterns from AI-generated content. The tool targets recurring signals such as em-dash overuse, filler phrases like 'in today's fast-paced digital landscape,' and words like 'delve' and 'robust' that research shows spiked sharply in frequency after ChatGPT's release. Its design draws on empirical sources, including a 2025 study by Kobak et al. finding 'delve' appeared at roughly 28 times its pre-ChatGPT rate in biomedical papers, and Pangram Labs data on inflated phrase frequencies. Alchemy works by flagging density of co-occurring tics rather than penalising any single word, and integrates with AI coding agents via a single init command. The tool is MIT-licensed and available on GitHub for anyone to read or adapt.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

How Freelance Developers Should Calculate a Fair Day Rate

Many developers transitioning to freelance work make the mistake of simply dividing their former salary by 260 working days, which fails to account for taxes, unpaid leave, and business expenses. A more accurate approach involves estimating actual billable days — roughly 210 per year after holidays and time off — then working backward from a realistic income target that covers all costs. Tools like PayCalcTools' free Freelance Day Rate Calculator can automate this process by factoring in country, holidays, and overheads to generate both a day rate and hourly rate. Industry benchmarks suggest US freelance developers can expect anywhere from $250 to over $1,000 per day depending on experience level. Experts also advise revisiting rates annually, since failing to adjust for inflation of 3–5% per year amounts to a gradual, silent pay cut.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

How Browsers Actually Pick a Font — and How Developers Can Detect It

A developer article on DEV Community explains that CSS's getComputedStyle method returns a font priority list, not the font actually rendered by the browser. The browser selects the first available font in the stack that contains a glyph for the character being displayed, a distinction that matters especially for Japanese text. Fonts like Hiragino, Yu Gothic, and Noto Sans JP differ visibly in weight and style, meaning a site designed on macOS can look different on Windows. Developers can detect the rendered font using canvas-based text measurement or the modern CSS Font Loading API via document.fonts. The author built a tool called Japanese Font Finder to automate this detection process.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Practical Golang Interview Prep Guide for Mid and Senior Engineers

A detailed preparation guide for Go programming interviews has been published, targeting mid-level and senior software engineers looking to sharpen their skills. The guide emphasizes that Go interviews go beyond syntax, testing candidates on concurrency, memory management, error handling, and system design trade-offs. Mid-level engineers are advised to focus on language fundamentals, testing, and the standard library, while senior candidates are expected to also demonstrate knowledge of Go's runtime scheduler, memory model, and profiling. The guide is structured as both a study path before interviews and a quick reference between rounds, rather than a list of random trivia questions. Its core message is that strong candidates must be able to explain code behavior, write correct programs, and articulate design decisions clearly.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Self-Taught Developer Builds First Python Project: A Command-Line Number Guessing Game

A self-taught Python learner has completed their first end-to-end programming project, a command-line number guessing game built using core Python concepts. The game gives players five attempts to guess a randomly generated number between 1 and 100, with feedback provided after each guess. Key programming concepts applied include functions, while loops, exception handling, and input validation to prevent crashes from invalid entries. The developer noted that breaking code into reusable functions was a major lesson learned, and that handling invalid user input was the most challenging part of the build. The project is available on GitHub, with the developer planning to tackle more complex applications going forward.

0 comments Read more at DEV Community