Turva.dev launches free tool to validate llms.txt files for AI agents

·2 views

Turva.dev has released a free llms.txt validator at turva.dev/llms-txt-validator, allowing users to check whether a website's llms.txt file meets the expected format. The tool fetches the file from a given domain and runs a series of checks, reporting each as pass, warn, or fail, with no data stored and no account required. The llms.txt format provides AI agents with a structured map of a website, reducing the cost of crawling by using a small, markdown-based file with a title, summary, and grouped links. The validator also supports a JSON API, enabling developers to integrate it into scripts or agent pipelines via a simple curl request. A self-referential bug was discovered and fixed during initial deployment, and the tool has since been verified against the llmstxt.org reference file.

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 Unplanned Work Derails Engineering Sprints and Strategies to Manage It

Unplanned mid-sprint requests — from urgent bug fixes to ad-hoc stakeholder demands — are a leading cause of missed deadlines and developer burnout in software teams. The core problem is not just lost time but the productivity cost of context switching, where a brief interruption can consume far more than the task itself requires. Experts suggest designating one team member per sprint as a dedicated handler for urgent issues, shielding the rest of the team from disruption. Teams are also advised to reserve a 20% capacity buffer for unplanned work and to log every request exceeding 15 minutes as a formal ticket for visibility. Without tracking how much unplanned work a team regularly absorbs, sprint planning will consistently over-commit available capacity.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Developer Automates Weekly Product Data Tracking Using n8n Scraper Workflow

A developer writing for DEV Community built an automated web scraping workflow using n8n to eliminate three to five hours of manual product data collection each week. The workflow uses a scheduled trigger, HTTP Request nodes, and CSS-selector-based HTML extraction to pull product names, prices, and availability from target pages every morning. Extracted data is automatically logged into Google Sheets with timestamps, while a conditional node sends Slack alerts only when values change between runs. The developer chose n8n over alternatives like Zapier, Make, and custom Python scripts for its open-source, self-hostable model and visual workflow editor. The article also emphasizes ethical scraping practices, including checking robots.txt, setting request delays, and mimicking standard browser behavior.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

How a Structured Review Framework Replaced 'Looks Good to Me' With Real Verdicts

A software engineer reflects on years of approving pull requests based on intuition rather than systematic requirement verification. Using a structured review phase that assigns verdicts — covered, partial, or missing — to each requirement, they discovered that working code does not always mean a requirement is fully satisfied. In one case, a notification scheduler routed calls through a queue in a way the original requirement never anticipated, prompting a 'pass-with-risk' recommendation rather than a simple approve or reject. The engineer argues that 'pass-with-risk' is not a failure verdict but a transparent handoff that names known uncertainties for downstream testing. The experience led them to conclude that traditional code review often mistakes a general impression of correctness for an actual evidence-based verdict.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

FlutterFlow Lets Vibe Coders Build Native Mobile Apps Without Writing Code

FlutterFlow is a visual, low-code platform founded by ex-Google engineers that lets users build native mobile applications through a drag-and-drop interface without writing code directly. Unlike AI vibe-coding tools such as Cursor or Claude, which generate code via chat prompts, FlutterFlow automatically produces clean Flutter code in the background as users design visually. A developer recently attended a FlutterFlow workshop and used the platform to build a Dev.to-style mobile app, highlighting how it lowers the barrier for non-coders to ship real apps. The platform supports deployment across phones, tablets, and computers, and gives developers full ownership of the generated Flutter codebase. With mobile development traditionally requiring large teams and long timelines, FlutterFlow is being positioned as a faster, more accessible alternative for builders looking to move from web to native apps.

0 comments Read more at DEV Community