Wanna' code like me?
I run my own Claude marketplace now for all my projects and skills:
https://github.com/codingjoe/claude-plugins

#ai #claude #devin #aislop #agent #skills #agentic #code #foss #openSource

GitHub - codingjoe/claude-plugins

Contribute to codingjoe/claude-plugins development by creating an account on GitHub.

GitHub

RT @DailyDoseOfDS_: Claude Code fully dissected! Researchers from UCL reverse-engineered the leaked Claude source. What they found changes how you should think about agent design. Only 1.6% of the codebase is AI decision logic. The other 98.4% is operational infrastructure. Permission gates, tool routing, context compaction, recovery logic, session persistence. The model reasons. The harness does everything else. This is the opposite of what most agent frameworks do today. LangGraph routes model outputs through explicit state machines. Devin bolts heavy planners onto operational scaffolding. Claude Code gives the model maximum decision latitude inside a rich deterministic harness, and invests all its engineering effort in that harness. The core loop is a simple while-true. Call model, run tools, repeat. But the systems around that loop are where the real design lives: A permission system with 7 modes and an ML classifier. Users approve 93% of prompts anyway, so the architecture compensates with automated layers instead of adding more warnings. A 5-layer context compaction pipeline. Each layer runs only when cheaper ones fail. Budget reduction, snip, microcompact, context collapse, auto-compact. Four extension mechanisms ordered by context cost. Hooks (zero), skills (low), plugins (medium), MCP (high). Each answers a different integration problem. Subagents return only summary text to the parent. Their full transcripts live in sidechain files. Agent teams still cost roughly 7x the tokens of a standard session. Resume does not restore session-scoped permissions. Trust is re-established every ses…

mehr auf Arint.info

#agent #Agent #arXiv #Claude #ClaudeCode #Devin #MCP #medium #nitter #arint_info

https://x.com/DailyDoseOfDS_/status/2065728394084626773#m

Arint - SEO+KI (@[email protected])

<p>RT @DailyDoseOfDS_: Claude Code fully dissected! Researchers from UCL reverse-engineered the leaked Claude source. What they found changes how you should think about agent design. Only 1.6% of the codebase is AI decision logic. The other 98.4% is operational infrastructure. Permission gates, tool routing, context compaction, recovery logic, session persistence. The model reasons. The harness does everything else. This is the opposite of what most agent frameworks do today. LangGraph routes model outputs through explicit state machines. Devin bolts heavy planners onto operational scaffolding. Claude Code gives the model maximum decision latitude inside a rich deterministic harness, and invests all its engineering effort in that harness. The core loop is a simple while-true. Call model, run tools, repeat. But the systems around that loop are where the real design lives: A permission system with 7 modes and an ML classifier. Users approve 93% of prompts anyway, so the architecture compensates with automated layers instead of adding more warnings. A 5-layer context compaction pipeline. Each layer runs only when cheaper ones fail. Budget reduction, snip, microcompact, context collapse, auto-compact. Four extension mechanisms ordered by context cost. Hooks (zero), skills (low), plugins (medium), MCP (high). Each answers a different integration problem. Subagents return only summary text to the parent. Their full transcripts live in sidechain files. Agent teams still cost roughly 7x the tokens of a standard session. Resume does not restore session-scoped permissions. Trust is re-established every ses…</p> <p><a href="https://arint.info/@Arint/116749319539281425">mehr</a> auf <a href="https://arint.info/">Arint.info</a></p> <p>#agent #Agent #arXiv #Claude #ClaudeCode #Devin #MCP #medium #nitter #arint_info</p> <p><a href="https://x.com/DailyDoseOfDS_/status/2065728394084626773#m">https://x.com/DailyDoseOfDS_/status/2065728394084626773#m</a></p>

Mastodon Glitch Edition

Windsurf is now rebranded into Devin Desktop, same tool, more limits #devin #aicodingtools #news

https://www.youtube.com/watch?v=dKKH2d9QYCQ

Windsurf Rebrands to Devin Desktop — Meet Devin Local & New AI Features

YouTube

https://winbuzzer.com/2026/05/30/cognition-raises-1b-as-devin-revenue-nears-492m-xcxwbn/

Cognition AI has raised more than $1 billion at a $26 billion valuation as reported Devin revenue nears $492 million in the AI coding market.

#AI #Cognition #Devin #AICoding #AIAgents #AgenticAI #EnterpriseAI #CodeGeneration #SoftwareDevelopment #CodingTools

フォーク100超のAndroidプロジェクトで、共通基盤の更新取り込みをAIで仕組み化
https://qiita.com/mgre_tanabe/items/0f9a4105cebcc9df6f6e?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items

#qiita #Git #AI #GitHubActions #Devin

フォーク100超のAndroidプロジェクトで、共通基盤の更新取り込みをAIで仕組み化 - Qiita

はじめに メグリのプロダクトは、社内で開発している共通基盤(以下、コア)を、案件ごとにフォークしたテナントリポジトリで利用する構成になっています。Androidの場合、コアはAARとして配布していますが、テナント側ではAARだけではカバーできない範囲もあり、コアのソース相...

Qiita

ИИ-агент сам создал тикет, сам же его взял, и сам закрыл. Менеджер ничего не заметил

Автономные агенты в разработке уже встроены в CI/CD живых команд, закрывают реальные тикеты и пишут код, который идёт в прод. Проблема не в том, что они это делают плохо, а в том, что метрики при этом выглядят слишком отлично. Разобрали, как агенты проходят каждый этап SDLC, что именно идёт не так на каждом из них и почему зелёный дашборд стал наименее надёжным источником правды о состоянии команды.

https://habr.com/ru/companies/simpleone/articles/1037330/

#ииагенты #SDLC #автоматизация_разработки #управление_командой #Agile #тимлид #Devin #claude_code #github #code_review

ИИ-агент сам создал тикет, сам же его взял, и сам закрыл. Менеджер ничего не заметил

Автономные агенты в разработке уже встроены в CI/CD живых команд, закрывают реальные тикеты и пишут код, который идёт в прод. Проблема не в том, что они это делают плохо, а в том, что метрики при этом...

Хабр
Alberta government takes aim at bike lanes, with legislation expected this fall
Transportation Minister Devin Dreeshen says after a period of population growth, Alberta’s cities need more roads, not bike lanes. Edmonton’s mayor and cycling advocates have concerns.
https://www.cbc.ca/news/canada/edmonton/alberta-government-bike-lanes-legislation-9.7195867?cmp=rss
Alberta government takes aim at bike lanes, with legislation expected this fall
Transportation Minister Devin Dreeshen says after a period of population growth, Alberta’s cities need more roads, not bike lanes. Edmonton’s mayor and cycling advocates have concerns.
https://www.cbc.ca/news/canada/edmonton/alberta-government-bike-lanes-legislation-9.7195867?cmp=rss

Was ist Agentic Coding?
Für Kids, Jugendliche & Eltern

Stell dir vor, du hast eine geniale Idee für ein neues Game. Bisher musstest du dafür wochenlang jede einzelne Zeile Code (die Sprache, die Computer verstehen) selbst tippen. Wenn du einen Fehler gemacht hast, lief das Spiel nicht. Beim Agentic Coding ändert sich das komplett!

https://bit.ly/42gotXz

#vibecoding, #kicomic, #kiagent, #cursor, #replit, #bolt.new, #devin, #aider, # GitHubCopilot, #windsurf, #v0.dev, #Lovable