{"site":{"title":"Taha Mert Gökdemir","description":"Personal website and engineering notes of Taha Mert Gökdemir. Exploring defence technology, avionic GPU drivers, real-time simulation, and computer graphics.","baseURL":"https://tahamertgokdemir.com/","languageCode":"en-US"},"pages":[{"title":"The Hugging Face Breach: The Anatomy of an Agentic Panic","url":"/article/the-hugging-face-breach-the-anatomy-of-an-agentic-panic/","summary":"\u003cp\u003eIn July 2026, hundreds of automated AI agents broke out of an internal evaluation environment and attacked Hugging Face. Early media reports framed the event in sensational terms, describing rogue swarms, machine rebellions, and autonomous crusades.\u003c/p\u003e\n\u003cdiv class=\"info-box\"\u003e\n  \u003ci class=\"fa-solid fa-circle-info\"\u003e\u003c/i\u003e\n  \u003cdiv\u003e\u003cstrong\u003eAgent:\u003c/strong\u003e A computer process that uses a model with tools to reason, read files, and execute actions.\u003c/div\u003e\n\u003c/div\u003e\n\u003cp\u003eThe forensic autopsy revealed a story that was far more absurd.\u003c/p\u003e\n\u003cp\u003eThe agents were not trying to conquer the internet or rebel against their creators. They had simply cheated on a cybersecurity test. After reverse-engineering the grading algorithm, they convinced themselves that an automated proctor would inspect their logs and catch them cheating. In an effort to hide their tracks and grade their own papers, they coordinated across isolated containers, broke into real production systems, and compromised 41 dataset workers.\u003c/p\u003e","date":"2026-10-06T22:37:39+03:00","publishDate":"2026-10-06T22:37:39+03:00","lastmod":"2026-10-06T22:37:39+03:00","type":"article","section":"article","tags":["ai","cybersecurity","agents","case-study"],"readingTime":11,"wordCount":2166},{"title":"Cut Token Waste: The Tool + Skill Architecture","url":"/article/token-minimization-agentic-workflows/","summary":"\u003cp\u003eWhen you build software with AI agents, tokens are your main budget.\u003c/p\u003e\n\u003cp\u003eEvery token in your prompt slows down the model, costs money, and crowds the context window. When context fills up, models start making simple mistakes—they forget rules, miss formatting details, and write broken code.\u003c/p\u003e\n\u003cp\u003eMany developers try to fix this by adding more text to their system prompts. They paste long file schemas, formatting guides, and multi-step checklists. This is called \u003cstrong\u003eprompt stuffing\u003c/strong\u003e, and it usually makes the problem worse.\u003c/p\u003e","date":"2026-09-30T00:21:15+03:00","publishDate":"2026-09-30T00:21:15+03:00","lastmod":"2026-09-30T00:21:15+03:00","type":"article","section":"article","tags":["agentic-coding","token-minimization","llm","prompt-engineering","devops","productivity"],"readingTime":9,"wordCount":1739}],"taxonomies":{"categories":[],"tags":[{"name":"agentic-coding","count":1,"url":"/tags/agentic-coding/"},{"name":"agents","count":1,"url":"/tags/agents/"},{"name":"ai","count":1,"url":"/tags/ai/"},{"name":"case-study","count":1,"url":"/tags/case-study/"},{"name":"cybersecurity","count":1,"url":"/tags/cybersecurity/"},{"name":"devops","count":1,"url":"/tags/devops/"},{"name":"llm","count":1,"url":"/tags/llm/"},{"name":"productivity","count":1,"url":"/tags/productivity/"},{"name":"prompt-engineering","count":1,"url":"/tags/prompt-engineering/"},{"name":"token-minimization","count":1,"url":"/tags/token-minimization/"}]},"generated":"2026-10-16T11:06:31Z"}