Developer Uses AI to Run 55 App Tests in a Week, Uncovers Key Bugs

·1 views

A developer shared their experience using AI to repeatedly test their web application over the course of a week. The AI was run through 55 testing cycles, surfacing a range of bugs and vulnerabilities in the app. Notably, some of the issues discovered were caused by the developer themselves, not just the AI. The experiment highlighted both the strengths and limitations of AI-driven testing in modern web development. The findings were shared on DEV Community as a six-minute read tagged under testing, AI, and DevOps.

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)

Developer documents three fixes needed to run gpt-oss-120b via Anthropic SDK on Sakura AI

A developer testing Sakura Internet's free AI Engine API encountered three consecutive roadblocks while connecting to the gpt-oss-120b model through Anthropic's Python SDK. The free tier, offered by Japanese cloud provider Sakura Internet, allows 3,000 chat completions per month at no charge, and the credit card requested at sign-up serves only as identity verification rather than a billing trigger. A 401 authentication error arose because Anthropic's SDK sends an x-api-key header by default, while Sakura's endpoint expects Bearer token auth, requiring the use of auth_token instead of api_key. A third issue emerged when indexing the response directly at content[0] failed due to a ThinkingBlock object appearing before the text block, which is resolved by filtering response content by type. All three issues have straightforward one-line fixes once their root causes are understood.

0 comments Read more at DEV Community

ProgrammingHacker News ·

Developer Builds Digital Twin Replays for All 64 World Cup Matches

A developer has created watchable digital twins covering all 64 games of the FIFA World Cup. The project involves building data-driven simulations that replicate real match events in a viewable format. Details of the technical approach are documented on the developer's personal blog. The work was shared on Hacker News, where it attracted early attention from the tech community.

0 comments Read more at Hacker News

ProgrammingDEV Community ·

Developer releases 'quota', a terminal tool to monitor AI coding tool usage limits

A developer has built an open-source terminal utility called 'quota' that displays remaining usage and reset times for AI coding tools like Claude Code and Codex. The tool reads provider data directly from local files such as ~/.claude and ~/.codex, requiring no API calls, accounts, or telemetry. Users can run it as a single command, set it to auto-refresh in watch mode, or pipe its JSON output into scripts. The project was motivated by the frustration of hitting rate limits unexpectedly mid-task, a pain point that drew significant discussion on Hacker News. The tool is available via npm and planned features include shell prompt integration and CI pipeline quota checks.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Developer builds open-source CLI tool to track AI API spending by project and model

A developer created AICostTracker, a local command-line tool that logs AI API usage and breaks down costs by project and model, after growing frustrated with opaque billing dashboards from providers like OpenAI. The tool runs entirely on the user's machine, storing data in a local JSONL file rather than a cloud service. It supports pricing for major AI providers including OpenAI, Anthropic, and Google, as well as local Ollama models. Users manually log each API call with details such as token counts and project name, then run a summary command to view spending patterns. The tool is available as an npm package on GitHub, with automatic API call interception planned as a future feature.

0 comments Read more at DEV Community