SaaSの権限管理設計を完全に理解した
datacurve-ai/deep-swe: Measuring frontier coding agents on original, long-horizon engineering tasks
Measuring frontier coding agents on original, long-horizon engineering tasks - datacurve-ai/deep-swe
DeepSWE
DeepSWE measures frontier coding agents on original, long-horizon software engineering tasks.
サプライチェーン攻撃に怯えないために設定した5つのこと - Plan 9とGo言語のブログ
ここ数ヶ月ほどサプライチェーン攻撃の被害が続いています。手元では各種パッケージマネージャを使いますし、AIエージェントやプラグイン機構のあるアプリケーションを使っているので、攻撃の影響を受けないとは言い切れません。サプライチェーン攻撃に対して完全な対策はおそらく不可能と思われるため、被害を受けたとしても影響を小さくする…
Turning stream of consciousness into real action | Google Cloud Blog
Discover how Doist is using gen AI and voice-to-text models to transform stray thoughts into practical to-do lists and more productivity features.
A new Register Allocator for ZJIT
We recently landed a new register allocator in ZJIT. Learn about lifetimes, interference graphs, and linear scan — and why a global allocator unlocks features like method inlining.
【GitHub Actions】actions/checkout には persist-credentials: false を設定するべき
itayinbarr/little-coder: A coding agent optimized to smaller LLMs
A coding agent optimized to smaller LLMs. Contribute to itayinbarr/little-coder development by creating an account on GitHub.
Doorman11991/smallcode: AI coding agent optimized for small LLMs. 87% benchmark with 4B-active model.
AI coding agent optimized for small LLMs. 87% benchmark with 4B-active model. - Doorman11991/smallcode
Claude Code の Dynamic Workflow を試してみた
Claude Code v2.1.154 で Dynamic Workflow と呼ばれる機能が追加されました。Dynamic Workflow は数時間から数日かかるような大規模な作業を実行するために設計されています。ワークフローは JavaScript で定義され、複数のサブエージェントをオーケストレーションすることができます。この記事では Claude Code の Dynamic Workflow を試してみた様子を紹介します。
retlehs/quien: A better whois and domain intelligence toolkit
A better whois and domain intelligence toolkit
colbymchenry/codegraph: Pre-indexed code knowledge graph for Claude Code, Codex, Gemini, Cursor, OpenCode, AntiGravity, Kiro, and Hermes Agent — fewer tokens, fewer tool calls, 100% local
Pre-indexed code knowledge graph for Claude Code, Codex, Gemini, Cursor, OpenCode, AntiGravity, Kiro, and Hermes Agent — fewer tokens, fewer tool calls, 100% local - colbymchenry/codegraph
類似画像検索モデルの開発ノウハウ
『【LINEヤフー データサイエンス Meetup】Kaggle Grandmasterが教える開発と解法&コンペ主催企業が語る舞台裏 他』の登壇資料です。
Everything That Went Wrong With Claude
A satirical timeline of Claude and Anthropic controversies, outages, bans, legal disputes, policy fights, and quality complaints.
Opus 4.8 benchmark results for AI code review and code generation
Opus 4.8 is the best model we have used for long-horizon agentic coding and code generation, and it holds its own on code review out of the box.
AIコードレビューとコード生成におけるOpus 4.8のベンチマーク結果 | CodeRabbit | 無料でお試し
Opus 4.8は、これまで使ってきたモデルの中で、長期的なエージェント型コーディングとコード生成に最も優れています。また、コードレビューにおいても、素の状態で十分に健闘します。
fs-safe — Safe filesystem primitives for Node.js
Capability-style filesystem roots for Node.js apps that handle untrusted relative paths.
openclaw/fs-safe: Race-resistant root-bounded filesystem primitives for Node.js.
Race-resistant root-bounded filesystem primitives for Node.js. - openclaw/fs-safe
openclaw/proxyline: Process-global proxy routing for Node.js
Process-global proxy routing for Node.js. Contribute to openclaw/proxyline development by creating an account on GitHub.
proxyline — Process-global proxy routing for Node.js
Process-global proxy routing for Node.js. One install routes node:http, node:https, undici/fetch, plus WebSocket and HTTP CONNECT helpers through a single explicit policy.
Michaelliv/pi-dynamic-workflows
Contribute to Michaelliv/pi-dynamic-workflows development by creating an account on GitHub.
SkillOpt | Executive Strategy for Self-Evolving Agent Skills
Project webpage for SkillOpt, a text-space optimizer that trains reusable natural-language skills for frozen language agents.
microsoft/SkillOpt: SkillOpt is a text-space optimizer that trains reusable natural-language skills for frozen LLM agents through trajectory-driven edits, validation-gated updates, and deployable best_skill.md artifacts.
SkillOpt is a text-space optimizer that trains reusable natural-language skills for frozen LLM agents through trajectory-driven edits, validation-gated updates, and deployable best_skill.md artifac...
evo-hq/evo: turns your codebase into an autoresearch loop — discovers what to measure, instruments the benchmark, then runs tree search with parallel subagents.
turns your codebase into an autoresearch loop — discovers what to measure, instruments the benchmark, then runs tree search with parallel subagents. - evo-hq/evo
How we contain Claude across products \ Anthropic
Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.
AI code reviews for complex codebases | cubic | Try free now
cubic is an Al code review platform that helps teams like Cal.com and n8n catch bugs and merge pull requests faster 28% faster.
usmanyunusov/nano-staged: Tiny tool to run commands for modified, staged, and committed files in a GIT repository.
Tiny tool to run commands for modified, staged, and committed files in a GIT repository. - usmanyunusov/nano-staged
utsengar/htmlbin-cli
Contribute to utsengar/htmlbin-cli development by creating an account on GitHub.
htmlbin — agent-first HTML hosting
API for agents to share HTML.
paraschopra/make-pages-interactive
Contribute to paraschopra/make-pages-interactive development by creating an account on GitHub.
