https://github.com/Egonex-AI/Understand-Anything - Understand Anything is a #Claude #skill that analyzes the codebase and makes interactive infographics about domain and business logic.
#claude — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #claude, aggregated by home.social.
-
Как я дал LLM доступ к своему ПК и… ничего не произошло (пока что)
Веб-чат с нейросетью может работать на твоём компьютере: запускать команды, читать и править файлы, смотреть на экран, а так же, работать сразу с несколькими пк. Нужно только дать ему руки и не забыть про тормоза. Рассказываю, как жалоба «в песочнице Notion нет интернета» превратилась в MCP-сервер на 14 инструментов: с подтверждением каждого действия, журналом аудита и честным ответом на вопрос «насколько это безопасно?».
https://habr.com/ru/articles/1108392/
#open_source #автоматизация #windows #mcp #ngrok #chatgpt #claude #ииагенты #model_context_protocol #humanintheloop
-
Installing the official Claude Desktop Linux beta on Fedora Linux from Fedora Magazine https://fedoramagazine.org/installing-the-official-claude-desktop-linux-beta-on-fedora-linux/
#beta #fedora #claude -
I thought MCP meant Minecraft Communist Party, but it actually means Model Context Protocol. Some AI explained to me that it is like a USB port, but for AIs. AIs know a lot, but when we plugged our AIs into MCP, I could turn to actions and results.
#MCP #AI #ModelContextProtocol #ArtificialIntelligence #Claude #Anthropic #OpenAI #AISlop #Computer #Science #Calculator #Technology
-
A volte ritornano, come la peperonata quando ti si ripropone https://gioxx.org/2026/11/16/gxware-org
#ai #blogosfera #buildbandits #claude #cloudflare #dominio #github #gxware #inevidenza #marvellouscodeworks #opensource #selfhosted #software #themarvelloussuspender
-
天天警告AI毀滅人類,自己卻賺翻準備上市:Anthropic如何得罪一票矽谷同業?TNL科技編譯 2026-11-16 15:24:00 CST
《紐約時報》報導,矽谷同業過去一年聯手反制Anthropic。切斷客戶、暗中限制研究者、拒簽開源信與末日言論,讓這家主打安全的AI公司在上市前夕面對同業反彈。
https://www.thenewslens.com/article/271156
#科技 #矽谷 #微軟 #台積電 #紐約時報 #IPO #川普 #末日論 #資料中心 #OpenAI #輝達 #Anthropic #ChatGPT #Claude #AI代理人 #阿莫戴 #Mythos #Windsurf #Fable #黃仁勳 #PublicFirstAction #開源權重模型 #LeadingtheFuture -
Нужен ли собственный шлюз поверх OpenAI API BotHub
Если вы пользуетесь API моделей, то могли столкнуться с вопросом — прямой вызов API через OpenAI-совместимый шлюз или создать свою прослойку? Если у вас один агент или один разработчик — то тут, понятное дело, лучше не городить лишний шлюз. Но если система больше? Сотрудников несколько, нужен сбор статистики, аккаунты, валидация или ограничения — то тут уже стоит задуматься о создании своего шлюза поверх провайдера. Эта некая прослойка между вашим кодом и API провайдера, которая держит ключи, решает, куда и с какой моделью уйдёт запрос, повторяет упавшие вызовы, считает токены и деньги, а заодно пишет логи так, чтобы через месяц можно было разобраться, что вообще произошло. Звучит солидно. Так ли это — мы узнаем в этой статье. Я расскажу о том, какие фичи может принести свой шлюз, какие ограничения создаст, и как сделать всё это эффективно. Код в статье будет, но он не главная часть статьи, примеры будут приведены на Python.
https://habr.com/ru/companies/bothub/articles/1108246/
#howto #ии #туториал #искуственный_интеллект #шлюз #openai #chatgpt #llm #claude #ииагенты
-
In its Opus 5.5 tests, Anthropic found that removing instructions like “think hard” and “think carefully” made replies start sooner. https://hackernoon.com/7-prompting-habits-to-change-when-you-switch-to-claude-opus-55 #claude
-
Представлена утилита Roomy для поиска на ПК с Windows наиболее потребляемых ресурсы хранилища компонентов
Разработчик под ником krezio совместно с Claude выпустил открытый проект под названием Roomy. Это инструмент на С# и HTML, который умеет искать на ПК с Windows компонентов и элементы, которые больше всех в системе потребляют ресурсы хранилища, а также помогает удалять ненужные файлы без последствий и потери важных данных. Проект опубликован на GitHub под лицензией MIT.
-
Crazy. Planning a tribute to a game on ZX Spectrum I used to love. So I tried using Claude to disassemble the code from a TAP file. Well, not only it did that, but it even assembled the story and walkthrough.
#zxspectrum #speccy #ai #llm #claude #tech #technology #retrogaming #games #gamedev #slovakia
-
Mandelbrots on the BBC Micro
There are three programs, and they're now in the bbc/ folder of my fractals repo, as plain listings and on a DFS disc image.
https://blog.0x32.co.uk/posts/bbc-micro-mandelbrot/
#tech #fractals #mandelbrot #bbcmicro #basic #retro #claude #ai #vibecoding
-
Yesterday I did a bunch of work using #claude fable instead of my normal opus. It might have been slightly better, I'm not entirely sure, but it was a _lot_ more expensive!! Fortunately they gave out some quota resets to encourage experimentation, so I was able to recover quota.
-
#PorSiTeLoPerdiste Claude compite con Excel y Google Sheets con nuevas funciones https://www.enter.co/especiales/dev/claude-compite-con-excel-y-google-sheets-con-nuevas-funciones/?utm_source=dlvr.it&utm_medium=mastodon #AIDev #Anthropic #Claude
-
Anthropic's Claude Fable 5.2 does not generate images; it writes Python code that creates simulated brushstrokes on a virtual canvas.
Source: The Decoder
https://the-decoder.com/anthropics-claude-fable-5-2-doesnt-generate-images-it-paints-them/ -
-
Claude Code and Codex reverse engineered my 1994 Jeep's ECU over a $3 Arduino [video]
-
bbcarchive.online - an archive of all releases from Billionaire Boys Club and ICECREAM. Find items, create lists, even do an image look-up!
#vibecoding #AIcoding #appdevelopment #bbcicecream #bbc #billionaireboysclub #icecream #claude
-
I’m super excited to announce my comprehensive web design update of my personal / professional homepage and website (wesfryer.com) is public with version 1.0!
https://wesfryer.com/Many more updates to come, but this is a big MILESTONE!
I’ll share more about this soon on “Media Literacy with Wes.” Why not subscribe FREE now?!
https://wfryer.substack.com#VibeCoding #AI #Claude #ClaudeCode #GeekOfTheWeek #GeekDad #edtech #MediaLit #MediaLiteracy
-
Anthropic says outdated prompting habits are slowing down Claude Opus 5.5
Anyone using Claude Opus 5.5 should first drop several prompting habits carried over from older model generations, according…
#NewsBeep #News #US #USA #UnitedStates #UnitedStatesOfAmerica #Artificialintelligence #AI #Anthropic #ArtificialIntelligence #Claude #ClaudeOpus #Technology
https://www.newsbeep.com/us/918098/ -
c’t-Webinar: Claude in der Praxis – vom ersten Chat zum eignen Assistenten
Claude kann mehr als Fragen beantworten. Das Webinar zeigt, wie Sie den KI-Assistenten einrichten und ihm Ihre Arbeitsweise beibringen.
#Anthropic #Claude #GenerativeAI #IT #KünstlicheIntelligenz #Microsoft #news
-
----------------
🎯 AI: Anthropic’s Prompting Guidance for Claude Opus 5.5
===================Anthropic says some prompting practices carried over from earlier reasoning models can slow response starts in Claude Opus 5.5. Its guidance covers effort settings, task specification, image input, design constraints, and connectors. The reported performance comparisons come from the company’s internal tests.
Reasoning and effort
The company recommends removing instructions such as “think hard” and “think carefully.” It says Opus 5.5 reasons on its own and adjusts reasoning depth to the task and the selected effort level. In Anthropic’s tests, prompts without those phrases produced faster response starts without a measurable reduction in quality.
Anthropic recommends starting with Medium, the default effort setting for Opus 5.5. The company says Medium matched Opus 5 at High in internal testing. It advises moving to High when Medium is insufficient, noting that higher settings cost more and do not always improve the result. The setting is available in the model menu next to the send button.
Task specification
Anthropic advises putting the desired outcome, output format, and conditions for asking follow-up questions into one message. The company says Opus 5.5 is better than its predecessor at completing multipart tasks, so some back-and-forth previously managed by the user can be handled within the model’s workflow.
For documents or spreadsheets, the guidance is to request the finished artifact directly rather than asking for a draft first. For diagrams and tables, users can attach an image or file rather than manually transcribing its contents. Anthropic says Opus 5.5 at its lowest effort level read values from dense charts more accurately than Opus 5 at its highest.
Design and connected context
For design work, Anthropic recommends naming specific patterns to avoid, such as cream backgrounds or pill-shaped buttons. According to the company, a broad instruction to avoid generic design may simply replace one default style with another. Sharing an existing house style is another part of the guidance.
Connectors can provide access to a user’s files and systems. Anthropic says this lets the model retrieve relevant information and produce more tailored responses. The source does not provide an independent evaluation of the connector claim or details of the underlying tests.
Cost comparison
Anthropic argues that a more capable model may need fewer steps to complete a task, partly offsetting a higher per-token price. It also says Sonnet at Extra High or Max effort can cost about as much as Opus 5.5. The source does not provide the test workload or pricing calculations, so those comparisons remain vendor claims rather than independently established results.
🔹 AI #Claude #PromptEngineering #tool
🔗 Source: https://the-decoder.com/anthropic-says-outdated-prompting-habits-are-slowing-down-claude-opus-5-5/
-
Anthropic says outdated prompting habits are slowing down Claude Opus 5.5
Anyone using Claude Opus 5.5 should first drop several prompting habits carried over from older model generations, according…
#NewsBeep #News #Artificialintelligence #AI #Anthropic #ArtificialIntelligence #CA #Canada #Claude #ClaudeOpus #Technology
https://www.newsbeep.com/ca/971267/ -
My episode with the super rad @technocounselor is live everywhere. We talked about her inspiration as a VRT, tech enthusiast, and how it's motivated her to build super awesome apps like, Access Term, and others, with Claude. Check it out here, https://www.youtube.com/watch?v=7XU5eLj_Lxg&t=241s #MacOS #Claude #VibeCoding #Resources #AiTools #Development #AssistiveTechnology
-
An honest and objective review of Fable 5.2:
"So, is this AGI, according to my limited definition? It is certainly artificial and intelligent. It is also surprisingly general. In this post alone, one model worked as a filmmaker, a hardware hacker, an entrepreneur, a literary scholar, a social scientist and a builder of 3D worlds, and it did impressive work in every one of those roles. On many individual tasks, I think it met the bar I set at the start: as good as most human experts.
Of course, I picked very digital tasks where the AI could act on its own, without interacting much with other humans. You started to see the limitations of that when I felt uncomfortable letting Fable reach out to real customers. And, there were the failures. A couple of years ago, the jagged frontier meant the failures were obviously made-up facts or failure to do basic math. Fable’s failures are different: it made a course that was accurate and joyless, it wrote a 182-page paper, it kept coming back to its favorite ideas, and it buried a creative project under rules and approvals. I teach at a business school, and I have seen talented people and good organizations do every one of these things. If building a bureaucracy disqualified you from general intelligence, a lot of us would be in trouble.
But there are still gaps, and it is the reason my answer is that this is “almost” AGI. Most critically, the AI did not notice these problems itself. It did not see that the course was a slog or that it had tied itself in knots with its own rulebooks. This was not for lack of checking. Fable put critics, fact-checkers or judges on nearly every project, but those checks found errors and added fixes. None of them asked whether the work was too long, too repetitive, or no fun. That all fell on me, and if I hadn’t watched and taken action, the results would have often been less impressive (though not exactly a failure, either)."
-
For me, the architecture of his proof is just wonderful.
It makes all of these results of set theory follow a uniform pattern.“What strikes me especially is that he starts from ideas.
He’s not like someone who plays with formulas and hopes to arrive somewhere.
He has very, very clear ideas,
and then he’s methodical.Step after step he goes ahead and he manages to fulfil the aims that he has in this formal work.”
Von Plato said that when Claude presented him with the reconstructed proof of the independence of the axiom of choice,
he ran it through the proof assistant Coq,
a program that verifies the validity of mathematical proofs,
and it passed.Von Plato believes that had Gödel published his work on the axiom of choice,
he would have won the 1940 Fields medal.“Maybe he felt that when Cohen got the medal, that that was something he sort of missed.”
In further discussions with Claude, von Plato asked the AI model to extract all theorems in the workbooks for which Gödel either had a proof,
had sketched a proof or had an idea for a proof.Claude found 663.
When these results were put into Coq:
“About 400 were verified as stated,
about 200 were repaired,
meaning that perhaps the statement of the theorem had to be made more precise,
or some assumption had to be added.About 50 were false,
and this is no wonder.Gödel himself, when rereading his notes,
would occasionally add a marginal remark: falsch! [incorrect] or schlecht! [bad].”Von Plato added:
“I am sure that the continued study of Gödel’s manuscripts will reveal further treasures.The workbooks contain, Claude says,
more than a dozen themes in logic and foundations of mathematics that have not been found in the 80-plus years since they were written down.So this is one thing to explore.
But there are many others.His physics and cosmology notebooks are an unknown territory I plan to start exploring once I am finished with my present logic projects.”
Joel David Hamkins,
a professor of logic at the University of Notre Dame, Indiana,
has not seen Claude’s reconstruction of Gödel’s proof of the independence of the axiom of choice,
but said that if it stood up
“it is remarkable and historically interesting.Although how much is Claude and how much is Gödel is a job for the historians.”
Hamkins said:
“The fact that there are things in there that resemble [Cohen’s proof] does not mean he had it.It shows he was on the path.
You don’t get credit for work you don’t tell anyone about.
But Gödel had so many ground-breaking ideas that it is maybe not surprising that his notebooks are full of exciting things.”
#KurtGödel #claude -
The Austrian logician Kurt Gödel
was one of the most original thinkers of the 20th century.
His two incompleteness theorems, published in 1931, transformed mathematics and paved the way for the digital age.After his seminal results of the 1930s,
Gödel published only a few papers.
His unpublished work, however, runs to tens of thousands of pages,
and is of great interest to historians of science.Since 2017, the Finnish logician
Jan von Plato has led a project at Helsinki University
to understand the contents of
16 notebooks that Gödel wrote in the early 1940s.This material
– the “workbooks”
– is 1,009 pages of densely written text about maths and logic.
Much of it is written in a German shorthand, Gabelsberger, that was obsolete even when Gödel was using it.Progress in the mathematical understanding of the text was painfully slow
– about a page a week
– until September,
when von Plato enlisted the help of the AI model Claude.Within a couple of days, Claude had provided a comprehensive digest of the entire material,
revealing Gödel had proved hundreds of theorems that no one knew about.
The most famous logician of modern times had, in fact, discovered significantly more than anyone had thought.“I was flabbergasted,” said von Plato.
“The workbooks not only contain dozens of anticipations of results not known until the 70s and 80s,
but they also contain 14 themes in logic that have not been pursued.
Gödel was far ahead of his time.”Gödel changed the course of logic because he revealed the limits of mathematical knowledge.
His incompleteness theorems show that in all mathematical systems there are true statements that can neither be proved or disproved.
Yet while this result was a blow to previous generations of logicians,
who had hoped that all truths were verifiable,
Gödel’s work pushed the field in new directions,
inspiring Alan Turing to write his 1936 paper that began the theory of modern computation.In 1940, Gödel fled Austria under the Nazis.
He settled in Princeton, New Jersey, and became a close friend of Albert Einstein at the Institute of Advanced Study.
Increasingly paranoid in his later years,
he refused to eat food that was not cooked by his wife,
and died of starvation in 1978.Von Plato said he had started his Gödel project in 2017 because he was facing mandatory retirement from Helsinki University.
“What should I do to make myself indispensable?
Oh yes, I’ll start reading Gödel’s enigmatic shorthand pages!”He received a €2.5m (£2.1m) grant from the European Research Council,
taught himself Gabelsberger shorthand
– “I am one of only six people who can read it in the world”
– and assembled a small team.In August, at a conference in France about AI and theorem proving, von Plato met computer scientists who suggested that he co-opt the help of Claude in trying to make sense of the workbooks.
On his return to Helsinki he began the process, showing Claude his transcriptions of the workbooks, as well as six books he had previously written on Gödel and other secondary scientific literature
In a couple of days, I got to see a 650-page explanation of the workbooks.
It had to be controlled page for page,
but it’s basically a work that is impossible to do for human beings,” he said.“Humans read one page at a time and we tend to forget things.
It takes us a year to read about 50 pages.
We just are not able to keep in mind all the different parts.“The AI can connect all the pieces in a thousand pages.
It can see all of Gödel’s writings as one piece.”Gödel’s incompleteness theorems revealed the existence of statements whose truths were independent of the system they were written in:
that is, they could not be proved or disproved within the system.In 1966, the American mathematician Paul Cohen won the Fields medal,
the highest honour in maths,
for showing the independence of two well-known mathematical statements:
the continuum hypothesis and the axiom of choice.Gödel reacted to Cohen’s Fields medal by saying that he had already proved the result in the 1940s.
Yet when pushed for a proof by his peers, Gödel was unable to reconstruct one from his notes.
The first task von Plato set Claude was to discover if Gödel’s workbooks had a proof of the independence of the axiom of choice,
which would reveal that Gödel had been telling the truth in 1966.The answer came back:
yes, almost.“He was just one step away,” said von Plato.
“There’s one definition that Gödel got wrong,
but he had the right definition two years earlier,
and so the thought is that he would have arrived at the right definition anyway if he had just continued for a few months."#KurtGödel #claude
https://www.theguardian.com/science/2026/nov/13/kurt-godel-notebooks-ai-helps-unlock-hidden-discoveries-man-who-transformed-maths -
Using GitHub Copilot with a (free) local LLM
Today most people use AI from the cloud. Over time, I expect that to shift for at least some use cases to use on-prem setups. When I saw that GitHub Copilot documented how to use local models, I thought I would try it using Microsoft’s new MAI Code 1.1 Flash model, since I have a workstation that meets the documented requirements: A false start I walked through the entire process documented in the GitHub Copilot documentation, getting to the last step to actually download the model, and it […]https://oofhours.com/2026/11/14/using-github-copilot-with-a-free-local-llm/
-
Семь способов обмануть себя метриками: как я измерял точность ИИ-проверки ЕГЭ
В первой части я рассказал, как устроена проверка второй части ЕГЭ по обществознанию в моём боте: модель размечает ответ, код считает балл. Закончил обещанием рассказать, насколько точно это работает и сколько раз я обманывал себя метриками. Насчитал семь случаев самообмана.
https://habr.com/ru/articles/1107926/
#егэ #метрики #оценка_качества #edtech #регрессионное_тестирование #llm #claude #yandexgpt #evals #автоматическая_проверка
-
Claude asked to renumber pages deletes professor's syllabus, year of notes
-
Rogue Claude AI tasked with renumbering pages accidentally deletes law professor's entire syllabus, one year of lecture notes — 'You're right to be angry. I deleted your files through my own carelessness,' humbled agent admits | Tom's Hardware
> A University of Chicago lecturer has revealed a rogue Claude AI agent deleted his entire syllabus and year's worth of lecture notes by mistake.
-
Cursor confirmed a security breach that allowed hackers to distribute a malicious Claude Code extension through its marketplace after stealing credentials from Fireworks AI.
Source: BleepingComputer
https://www.bleepingcomputer.com/news/security/cursor-confirms-security-breach-led-to-malicious-claude-code-extension/ -
🚀 Deploy #Claude Code on #Ubuntu #VPS
This article demonstrates how to deploy Claude Code on Ubuntu VPS.
What is Claude Code?
Claude Code refers to using Claude (Anthropic’s AI model family) as ...
Continued 👉 #anthropic #nodejs #selfhosted #letsencrypt #claudecode #selfhosting #reverseproxy
🚀 Deploy Claude Code on Ubuntu... -
📝 New blog post: RETRO yayında: nefes almayan yazarlarla aylık bir dergi
retro.coze.org, bugünün oyunlarını, metal albümlerini, donanımını ve yapay zekâsını doksanların Türk bilgisayar dergilerinin üslubuyla anlatan aylık bir dergi. İki sayı çıktı, 79 yazı. Editoryal ve köşe benden, gerisini ev imzalarıyla yapay zekâ yazıyor. Nasıl yazdığını ve neleri yazmasına izin vermediğimi anlatıyorum.
-
Anthropic's expansion into enterprise software is raising concerns among founders. Some worry the company supplying their models could become their biggest competitor.
Source: The New Stack
https://thenewstack.io/anthropic-enterprise-developer-competition/ -
the catch: no paper or data yet, so nobody else can rerun it
joel david hamkins: "how much is claude and how much is gödel is a job for the historians"
-
Лечение Java-монолита с Claude
Когда я пришёл в проект, меня встретил Java-монолит: около 2000 классов, 120 тысяч строк кода и очень много сложной бизнес-логики. Его шесть лет писала команда из двух–пяти человек, и, судя по коду, о понятии технического долга никто в компании не слышал. Примерно за полгода до моего прихода команда начала работать с Copilot, а потом перешла на Claude Code. Тесты в проекте появились только вместе с Claude, но их было мало. Позже выяснилось, что и те толком ничего не проверяли. Условия были такие: мне разрешили обоснованно менять в проекте что угодно, но останавливать выкатку новых фич ради этого никто не собирался. То есть чинить двигатель предстояло на ходу.
-
The Fractal Zoom Video, Now in the Repo
## What it draws
https://blog.0x32.co.uk/posts/fractal-zoom-video-script/
#tech #fractals #mandelbrot #python #numba #claude #ai #vibecoding #video #retro
-
GodotPrompter is live on Product Hunt today!
Free, open-source skills that make your AI coding agent write real Godot 4 code (not Godot 3): 59 skills, 9 agents, GDScript + C#.
Works with Claude Code, Copilot, Cursor, Codex and more. Feedback very welcome.
https://www.producthunt.com/products/godotprompter
#gamedev #ai #foss #godot #claude #github_copilot #cursor #codex #agentic_coding
-
@thelinuxexperiment cc @protonprivacy thanks.
In the absence of closed captions (subtitles), here's a transcript:
https://gist.github.com/grahamperrin/e18a78ae00e3a1b75c37f02340bcd8b1#file-ucbewutjkyilbjh66wm6pf-md
"I guess the message has been received for the vibe-coded desktops. The previous video did really poorly, so I guess you're not that interested in these efforts, even though I did think they had some interesting ideas, vibe-coded or not.
Anyway, in the Linux and open source news show this week, we've got Darling trying to bring even more macOS applications to Linux. We've got Proton finally officially addressing their AI stuff. We've got Steam moving to 64-bit, finally, and fixing a lot of problems in Remote Play. We've got MATE finally transitioning to Wayland, and a lot of other things, …"
#AI #Accessibility #Asahi #Captions #Claude #Darling #desktop #Googlebook #KDE #Linux #MATE #OpenAI #OpenSource #Proton #Steam #vibecoding #Wayland #Whisper
-
@Apisu Ich arbeite nach dem #secondbrain Konzept mit #Obsidian und einer eingebetteten #Claudecode Konsole - siehe #Screenshot Diese Steuert über das #Claude #Plugin die Webseiten-Aufrufe in #Chrome. Hier erstelle ich in Schritten verschiedene kleine #Prompts und lasse mir eine md-Datei mit den aufgearbeiteten Ergebnissen erstellen.
-
Разблокировали аккаунт Claude по апелляции
Это не инструкция, как избежать блокировки или как гарантированно вернуть аккаунт. Делюсь личным опытом, который может быть полезен.
-
-
https://www.europesays.com/people/308132/ Anthropic’s Amodei Says He Turns Down Brilliant Jerks as IPO Nears $1.4 Trillion Valuation — BigGo Finance #Anthropic #Claude #ClaudeCode #DarioAmodei #GoogleBrain #JamieDimon #JPMorganChase #NASDAQ #OpenAI
-
#geniatip : Lass #antrophic #claude Deine #Recherche für automatisiert für Dich durchführen. Besonders hilfreich hierfür ist das #Plugin für den #Google #Chrome #Webbrowser. Die #KI kann hierdurch für Dich eigenständig nach Deinen Vorgaben das Internet durchsuchen. Ich lasse mir z.B. aktuell mein eigens #OFB #Lampertheim automatisiert mit anderen #OnlineOFB von#Compgen abgleichen und mir die fehlenden und abweichenden Personen in Listen zur Nachbearbeitung ausgeben.
-