Community posts

Claude Code mods are the biggest change to it this year. One line in Anthropic's own announcement matters more than all three demos: "They aren't sandboxed." A mod is TypeScript that sits in the middle of every event in your session. It can rewrite your prompt before the model sees it, block or rewrite a tool call, approve or deny a permission request, and redraw the UI. Through the mods API it can read files, run processes and make HTTP calls, with the same access Claude Code has on your machine. That's what makes them powerful. It's also what makes a bad one dangerous. Who's protected by default: - Team and Enterprise plans, or any machine with managed settings, load a built-in mod called sec-default first. It stops installed mods from doing risky things like overriding your permission deny rules. - Everyone else installing from a random repo has no such layer. You're trusting the author. The upside is huge. Blast Radius shows what rm -rf or a force push would delete before it runs. Token Weather shows how full your context is. Anthropic already ships /diff and AGENTS.md support as mods, and you can switch them off or replace them. So the takeaway: build your own mods, or ask Claude to write one. Read the code of anyone else's before you run /plugin install. Announcement: http://claude.com/blog/claude-code-mods Build your first one: http://claude.dev/blog/getting-started-with-claude-code-mods sec-default source: http://github.com/anthropics/claude-code/tree/main/mods

صدر ⁦Claude Code 2.1.287⁩ مع ⁦106⁩ تغييرات في واجهة سطر الأوامر. تتيح ⁦Claude Mods⁩ الآن للإضافات تعديل سلوك المساعد بشكل أعمق لتمكين ملحقات أكثر غنى وسير عمل مخصصة. تم إصلاح أمر ⁦rm⁩ الخطير الذي كان يفقد موجه التأكيد الدائم عند إعادة توجيه المخرجات. Claude Code 2.1.287 has been released with 106 CLI changes. Claude Mods now allow plugins to modify deeper assistant behavior for richer extensions and custom workflows. Fixed a dangerous rm command that was losing its always-ask prompt on output redirection.

Claude Codeが早くも「自分専用IDE」に近づいてきた印象です。 TypeScriptを数行書くだけで コードアシストのふるまいを変えたり UIをカスタマイズしたり 独自機能を差し込めるモッドが作れるようになった。 作ったモッドはプラグインとして配布できるので チームやコミュニティ単位で「仕事に最適化されたClaude」を共有できる。 公開されている例も面白い。 Token Weatherは コンテキストウインドウの使用状況を スパークライン付きで可視化してくれる。 Blast Radiusは rm -rfやgit reset –hardのような危険コマンドを検知して 実行前に影響範囲を見せてくれる。 Replay Theaterは Claudeが行ったファイル編集を記録して 一手ずつ巻き戻しながら確認できる。 どれも人間側の注意力や記憶力に頼っていた部分を 機械に肩代わりさせる設計になっている。 省エネ経営の視点で見ると 「優秀な開発者を増やす」のではなく 「普通の開発者が事故らず速く書ける環境を量産する」方向に AI IDEの競争軸が動き始めているように見える。 自社の開発フローに合わせたモッドを一つずつ積み上げていくと 同じ人員でもアウトプットが静かに変わっていきそうです。 ────────────────────── ▼ 補足情報 元ポストではClaude Code向けのモッド機能が紹介されている。モッドはTypeScriptで定義し、Claude Codeのプラグインとして読み込む形になっている。公式ブログのガイドではサンプルモッドがClaude Code Playgroundリポジトリに順次追加されるとされており、開発者はそれをベースに改造する想定になっている。 Token Weatherは対話ごとのトークン消費を視覚化するモッドで、長時間のセッションでコンテキストがあふれるリスクを把握しやすくなる。Blast Radiusはシェルコマンド実行前に対象ファイルやディレクトリを一覧してくれる仕組みで、ヒューマンエラー由来の障害低減が狙いと考えられる。Replay Theaterはセッション内の編集履歴を差分として追えるため、レビューやロールバックの手間の削減につながる設計になっている。

@ClaudeDevs claude code just became an editor. give it six months and someone ships a mod that turns it into vim

@ClaudeDevs The plugin model is the key here: customize behavior and UI without forking Claude Code. That makes agent workflows composable instead of brittle. Curious to see what developers build first.

Guys shipped a literal NUKE. 💀 I can now mod my Claude Code desktop app, wow.

Claude Code CLI 2.1.287 changelog: New features: • Added Claude Mods: plugins may now modify deeper behavior • Added You should know, an opt-in plugin where a side agent watches your back and flags things you or Claude might miss. Turn it on with /plugin enable cc-plugin-you-should-know@builtin (for first-party sessions with telemetry on) • Added an n:<text> filter to the agents view that matches session names and tasks; a filter now shows matches in collapsed sections and Enter opens the first match • Added prompt_text to the OpenTelemetry user_prompt event, a copy of prompt for backends that nest dotted keys; drop or mask it wherever you drop or mask prompt (anthropics/claude-code#70763) • Added URL prompts from MCP servers on the 2025-11-25 protocol, for example to sign in. If a server no longer connects after this update, add "bareElicitationCapability": true to its MCP config entry • Windows: Added a startup warning when denying the Bash tool also turns off the PowerShell tool, so Claude has no shell tool Fixes: • Fixed fast mode staying off in remote sessions owned by an agent with no user account, even when the organization allows it • Fixed Remote Control not receiving messages for minutes at a time when a reconnect request got no response; it now gives up after 30 seconds and retries • Fixed hooks configured with asyncRewake waking Claude over and over with "found issues" notifications when the hook's script file is missing; the broken hook is now reported once • Fixed tool heartbeats not reaching SDK hosts while the model's response stream was stalled with no data arriving • Fixed Bedrock and Vertex startup model checks ignoring an enforced availableModels list, which could collapse /model to one Opus row • Fixed the Claude in Chrome browser picker showing a JSON parse error when Chrome could not be reached • Fixed picking Fable in /model on a http://claude.ai login saving the current version's id, so your saved default now follows the newest Fable like Opus and Sonnet do • Fixed switching between Opus 5.5 and Sonnet 5.5 (/model, opusplan) rewriting earlier MCP tool announcements, which could drop earlier extended thinking • Fixed Amazon Bedrock Guardrails blocks that arrive mid-response ending the turn with an API error instead of the guardrail's message when the reply began with thinking • Fixed a dangerous rm (such as one on / or the home directory) losing its always-ask safeguard when the same command also redirected output to a ~ or wildcard path • Fixed claude -p and SDK sessions repeating a model fallback on every later message after the model was switched while a reply was running • Fixed a folder's CLAUDE.md being attached a second time after resuming a session or after a compaction • Fixed background sessions that could not be reopened from claude agents after the agent exited and removed the worktree the session was started in • Fixed /advisor pairing checks: Sonnet 5.5 can now advise Opus 4.7 and 4.8, and advisors the API would refuse are flagged up front instead of being silently dropped • Fixed Bash permission prompts showing internal parser names such as "Contains simple_expansion" instead of a plain explanation • Fixed a cause of fullscreen sessions on slow or busy machines exiting with "Claude Code exited after an unrecoverable interface error" while a scroll key was held in a long conversation • Fixed organization per-tool permission ceilings being silently dropped for an MCP tool named proto • Fixed Claude being told to page large MCP results saved as JSON with Read's offset and limit, which cannot split one long line • Fixed the commit attribution reminder being delivered inside a tool result after a compaction • Fixed screen reader mode leaving the cursor away from the typed text in search boxes (such as /resume and /permissions) and sign-in code fields • Fixed screen reader mode refusing Enter with nothing typed on /rewind's summarize options, whose added context is optional • Fixed screen reader mode showing a "Tab to amend" hint on approval prompts, where Tab does nothing • Fixed screen reader mode listing arrow keys that do nothing in /permissions and /mcp, and saying "Select with numbers" in empty menus or while a search box has the keys • Fixed screen reader mode leaving out the changed lines in file edit approval prompts and other diffs • Fixed screen reader mode sending the claude --teleport progress screen, and an MCP form field while it is being checked, to the screen reader again on every spinner frame • Fixed CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS not removing the structured-output format from session-title and prompt-hook requests, which Bedrock-backed gateways reject • Fixed screen reader mode leaving out the top lines of a second approval prompt, a changed /config row or the rejected-plan line when the previous screen was taller than the terminal window • Fixed --include-partial-messages sending a cut-short reply's message_stop late or never, so apps could show the reply as still in progress • Fixed claude agents sometimes not showing the permission prompt a background session is waiting on • Fixed /ultrareview giving advice about .git/info/attributes when the upload stops on a committed .gitattributes it cannot read, such as one saved as UTF-16 • Fixed claude remote-control failing to register behind an HTTP proxy with a misleading "Check your organization permissions" error (anthropics/claude-code#97352) • Fixed sandboxed Bash commands on Linux inheriting an open handle on the Claude Code executable • Fixed the running-tool dot and three spinners still moving with the "Reduce motion" setting on, and /rewind's confirm screen updating its "ago" time while you type a note • Fixed times in claude agents changing every second in screen reader mode; they now change at most every 10 seconds • Fixed a revoked http://claude.ai login showing a generic API Error: 401 instead of "OAuth token revoked"; in -p mode the error now starts with "Failed to authenticate" • Fixed /ultrareview upload refusals advising you to copy a variable named by a repository's settings file into your own user settings • Fixed --output-format stream-json and the SDK not streaming the turns of a context: fork skill run by typing /<skill> as the prompt, as they do for the Skill tool's fork • Fixed /feedback and /bug: the pre-filled GitHub issue no longer includes your recent error messages, and the confirmation screen now lists them as part of the report • Fixed claude plugin marketplace add --sparse and git-subdir plugin installs failing with "transport 'http' not allowed" when the repository is served over plain http • Fixed cloud sessions sometimes losing the earlier conversation when the session restarted while it was being compacted • Fixed a plugin reload that overlapped the startup --plugin-url download corrupting the session's cached copy of the plugin archive • Fixed /desktop quoting partial output when opening Claude Desktop timed out or printed too much output; the error now names the cause • Fixed an MCP connector tool call occasionally running twice, or the connector's calls failing until restart, when its server changed which MCP protocol version it supports • Fixed SessionStart hooks from synced plugins not running in new cloud sessions • Fixed the transcript's "N hooks ran" summary and the verbose debug log's matched-hooks count including Claude Code's internal callbacks, so one configured hook no longer shows as two • Fixed files Claude sends from cloud and Remote Control sessions failing when the upload finished just after the 30-second timeout; it now waits 35 seconds • Fixed repositories added mid-session in cloud and SDK sessions not loading their skills and plugins, and loading CLAUDE.md late, after Claude changed directory • Fixed PNG, JPEG and WebP images over 8,000 pixels on a side failing to send from a remote session; Claude now sends a scaled-down copy • Fixed messages sent from the Claude apps with 17 to 20 attached files delivering only the first 16 • Fixed headless sessions reporting an MCP server as needing authentication after one refused call, even though later calls succeed • macOS: Fixed Remote Control sessions started with claude remote-control stopping mid-turn when the Mac went to idle sleep • Windows: Fixed interactive claude hanging or crashing with "Raw mode is not supported" when its input is piped or redirected; it now says why and exits (use -p for piped input) Improvements: • Improved /config: settings that cycle show ‹ › and step both ways with ←/→, narrow terminals stack each value under its label, and PgUp/PgDn page the list • Improved plugin marketplace errors to say in plain words why a marketplace was ignored or refused, and what to do • Improved plugin listings to note when a plugin's dependencies were not installed, and updating a plugin now retries an install that did not finish • Improved the Claude apps gateway's error when Amazon Bedrock rejects a model ID: developers now see which model is unavailable, and the gateway log names the ID that was sent • Improved SDK sessions so a message sent with priority "now" no longer cancels a running web fetch or web search; it keeps loading in the background • Improved /memory: the left and right arrow keys now flip its on/off settings, such as Auto-memory • Improved /skill names typed mid-message: Claude is now told they are skills, including disable-model-invocation ones • Improved the contrast of the prompt input border in light themes and of the ❯ before your earlier messages • Improved delivery of files Claude sends from cloud sessions and Remote Control: an upload that fails on a timeout, a network error or a 502, 503 or 504 is now retried once • Improved what Claude says when a file cannot be sent for a reason that may be temporary: it now mentions that you can ask for the file again in a few minutes • Improved the prompt for a held message from another session to show the message between dashed lines, matching other permission prompts • Improved MCP and other tool permission prompts to show the tool call between dashed lines, matching file edit prompts • Improved MCP startup in headless mode: a remote server whose first connect fails transiently is now retried without waiting for the slowest server to finish connecting • Improved files Claude sends from a remote session: large files now stream from disk instead of being read into memory, and a file over the size limit is refused with the server's limit named • Improved the explanation Claude gives when the server refuses a file it sends from a Remote Control or cloud session, such as an oversized image • Improved handling of large MCP tool results: less memory, smaller session files, and no extra upload to count tokens for results far over the limit • Windows: Improved Bash tool speed by removing a subshell that ran before every command Security/safety changes: • Bedrock, Vertex, Mantle: Fixed model availability checks under CLAUDE_CODE_SKIP_*_AUTH sending a different Authorization header than real requests when ANTHROPIC_CUSTOM_HEADERS repeats it • Changed waiting permission prompts to show oldest first, so a new prompt no longer covers the one you're reading (prompts with a countdown still open on top) Other changes: • Self-hosted runner: Added a built-in gh api (REST only) for sessions that use Anthropic-managed git on macOS and Linux machines where the GitHub CLI is not installed • Changed a shell write through a repo-committed symlink onto a sensitive file or out of the working tree to name where it lands and wait for a person, on lines with a ~ target too • Changed Opus 4.7+ and Fable to use a 1M context window by default on Bedrock, Vertex, Foundry and the Claude apps gateway, with no [1m] suffix (CLAUDE_CODE_DISABLE_1M_CONTEXT=1 keeps 200K) • Changed replies from claude agents to arrive as queued messages; slash commands other than /stop sent while a turn is running now run when it ends • Changed whole-tool Bash allow rules and allowing hooks to prompt for, not run, shell writes to files Claude Code's file tools refuse outright (the Anthropic profile store, the host credentials file) • Changed right-click paste on Windows and Linux, and middle-click paste on Linux, to happen when the button is released; moving the pointer away before releasing cancels it • Changed MCP server alwaysLoad: false to defer all of that server's tools behind tool search • Changed screen reader mode to write new or changed lines without first pausing with the cursor at the start of the line; set CLAUDE_AX_PREPARK_MS=50 to restore the pause • Changed automatic model switches after a flagged message to keep your current effort level instead of the new model's default • [VSCode] Added "Run in background" to a running command or sub-agent, to move it to the background and keep working • [VSCode] Added the output of background shells and Monitors to their cards in the agent map • [VSCode] Fixed settings dialogs blaming a timeout when Claude Code's reply was too large to confirm a save • [VSCode] Fixed reopening a cloud session that the side bar already brought to this machine opening it again in a new tab; the side bar is shown instead • [VSCode] Fixed the side bar's Web tab not listing cloud sessions started after the window loaded; a failed load now says "Remote server is not connected" instead of "No web sessions yet" • [VSCode] Fixed a tab restored after a reload starting a second Claude process on a conversation the side bar already has open; it now shows the "still open somewhere else" notice • [VSCode] Fixed tool-row file links, session-list links and two hints showing in plain text • [VSCode] Fixed a background agent's still-running command showing as failed once the main turn ended • [VSCode] Fixed a user's own /usage or /context command opening the extension's dialog instead of running when picked from the command menu • [VSCode] Fixed file links in the plan preview tab doing nothing when clicked; they now open the file like links in chat replies • [VSCode] Fixed opening a tool's input or output in an editor tab failing with "Timeout waiting after 1000ms" on remote hosts such as WSL when the tab is slow to appear • [VSCode] Improved the Manage plugins dialog: a failed marketplace add, remove or refresh now says what went wrong • [VSCode] Changed the Claude in Chrome "Enabled by default" switch to also connect the editor's own sessions, which still ask before browser actions • [Cloud sessions] Fixed occasional failures to fetch from or push to GitHub when GitHub briefly refused a newly issued access token • [Claude Tag] Fixed Claude posting a failure warning, such as a spend limit notice, in a Slack thread when a background event like GitHub activity woke it and nobody was waiting on a reply • [Claude Tag] Fixed Claude Tag's spend limits page in admin settings leaving out recently created and private channels in organizations with many channels • [Claude Tag] Improved Claude's task list in long Slack threads: background work no longer reposts it as a new message on its own, so people following the thread aren't notified • [Code Review] Fixed finding comments and their "Why this was flagged" text stopping mid-sentence; they now end on a complete sentence • [Code Review] Fixed Code Review skipping a pull request after a new push when its review had failed twice on the previous commit; it now reviews the latest commit • [Code Review] Improved the failed-review card on a pull request whose conversation is locked: it now says the lock blocked the review and that nothing was posted or charged Source: https://github.com/anthropics/claude-code/blob/main/CHANGELOG.md#21287

Claude Code 2.1.287 has been released. 106 CLI changes Highlights: • Claude Mods allow plugins to modify deeper assistant behavior, enabling richer extensions and custom workflows • Fixed dangerous rm losing its always-ask prompt on output redirection, preventing accidental deletions Complete details in thread

@ClaudeDevs We got Mods in Claude Code before GTA VI.

Thursday afternoon AI FACTS roundup 🤖 01.10 - 🕵️ Chinese hackers impersonated officials and an Anthropic employee to spy on AI experts - 🏦 Barclays aims to roll out Anthropic's Claude to 50% of its developers by the end of 2026 - 📊 Adaption AI launched Invent a Dataset, outperforming GPT-5.6 by 17% in quality and 19% in diversity - 🎮 Universal Modder lets Claude Code automatically create mods for PC games across 12 engines - 💰 AWS ($AMZN) is raising reserved GPU prices by 15% for the fourth consecutive quarter. - ⚡ Amazon will buy 690 MW of nuclear energy from Constellation for 20 years in Maryland, USA - 🏛️ Judge Mehta dismissed Chegg's and Penske's antitrust lawsuits against Google - 💰 Photon raised $4.5 million in a seed round for its AI agent platform - 🛰️ Satlyt raised $8 million for software to run AI on satellites - 📈 Accenture shares rose 20% after its GAAP operating margin increased to 15.3%, driven by demand for AI deployments - ⚡ Volantis raised $88 million in Series A funding for optical memory chips - 📊 The AA-Video v2.0 benchmark evaluates video generators: Wan 3.0 leads, with MiniMax H3 at $4.80/min - 💵 Meta cut its 2025 tax bill by $3.9 billion by classifying AI data centers as research facilities - 📈 Legora surpassed $200 million in ARR in September 2026 - 🤝 Halluminate works with 4 of the top 5 AI labs in the USA - 💵 Nebius is acquiring Inferize for $100–150 million to reduce GPU idle time - 🇸🇮 Slovenia: .si domain registrations jumped 2100% after AI was dubbed “Super Intelligence” - 🏠 Airbnb rolled out AI search; Brian Chesky says it will evolve into an agent - 🤖 Robinhood launches Agent Apps, offering stock market data and Nasdaq sentiment from $10 for AI assistants - 🖥️ Runway unveiled Solaris, a model that generates no-code app interfaces as video - 🎬 HeyGen introduces a universal business video model based on MiniMax H3 - 👓 Startup Legato launched Legato Frames AI hearing glasses starting at $999 - 📦 Cloudflare launches the open beta of Artifacts for AI agents to manage Git repositories - 🪙 Blockworks MCP integrates data on 40,000 crypto assets with AI assistants - 🤖 Cloudflare unveils clef: its first in-house AI decision models with open-source weights - 🤝 Bolt acquires Dokai to add enterprise workflow support to its coding agents - ⚖️ Senators Hawley and Murphy propose a bill making AI companies criminally liable for hacks carried out by autonomous agents - 📊 Israeli AI and cybersecurity startups raised $2.77 billion in September - 🖼️ Mitsuba-ComfyUI-27B: a Qwen3.8-27B vision model compressed to 7.3 GB through ternary quantization - 🧠 A new AI tool can reconstruct images from brain scans - 🎙️ Fermion Research released the 164 MB Phonon-2 with an average WER of 5.21% - 🤖 Boston Dynamics equipped the Atlas robot with hands offering 13 degrees of freedom - 📊 Pangram: 31.1% of FineWeb tokens are AI-generated text - 🦴 A Princeton team released UniMate, an AI model that animates skeletons from text. - 🛡️ Unitary developed the unbiased-toxic-roberta model with an AUC score of 93.74 - 🔬 NVIDIA proposes PivotOPD, an on-policy distillation variant for multi-step language agents - 🛡️ Matthew Green warns of AI worms spreading between agents despite sandboxing - ⚡ Magnitude open-sourced an inference engine that runs up to 2x faster than llama.cpp - ⚡ SenseTime developed Looped-DiT, a 260M diffusion model that outperforms a 1.7B model with 4.9x less compute - 🛠️ ALGO ARTIS released yomiyasu, an open-source tool that makes AI-generated Japanese sound more natural - 🛠️ LiteLLM introduces Lens for optimizing AI agents based on traffic - 📈 A new RL proxy for OpenEnv boosts the LFM2.5-2.6B model's success rate from 42% to 54% - 🎮 OpenDLSS-NR reimplements NVIDIA's DLSS 5 Neural Rendering network in Vulkan Bookmark this — it's the fastest way to check what you missed today. Repost if someone on your timeline needs it. Twice a day, every day. 🤖 📊 AI MARKET PULSE 🧠 Intelligence frontier: 57.6 (Claude Opus 5.5 (Adaptive Reasoning, Max Effort, Default Fallback)) 🖥️ H100: $3.20/h ↑4.9% 7D ⚡ B200: $7.90/h ↓0.6% 7D 🔤 LLM tokens: $0.98/1M ↑2.2% 7D 🔥 Most used LLM: stealth/space-bunny-alpha Sources: Artificial Analysis · Ornn · Silicon Data · OpenRouter

We’ve used mods to build features like /diff and AGENTS.md support. We hope you find them as useful (and fun) as we do, and can't wait to see what you build!

We're adding pre-built sample mods to the Claude Code Playground repo you can riff off of, if you need some inspiration. Learn more in our getting started guide: https://claude.dev/blog/getting-started-with-claude-code-mods/

Keep mods to yourself, or submit them to the Claude directory if others might find them useful. Mods run with the same access to your machine as Claude Code itself, so make sure to only install mods from sources you trust. https://claude.ai/directory

Replay Theater records every file edit Claude makes in a turn. Run /replay to step through the diffs one at a time in a docked pane.

You can now mod Claude Code: - Change how it behaves - Customize the UI - Swap in your own features Write one with a few lines of TypeScript, or have Claude build it for you. Mods ship inside plugins, so you install them with /plugin in the CLI or desktop app. A few examples:

Token Weather adds a forecast of your context window: how full it is, plus a sparkline of your last 12 turns.

Blast Radius catches risky shell commands like rm -rf, git reset --hard, or a force push before they run. It shows what the command would touch in a side pane.

Claude Code v2.1.287 sürümünde yeni eklentiler ve düzeltmeler Claude Code v2.1.287, eklentilerin daha derin davranışları değiştirmesine ve isteğe bağlı You should know eklentisine olanak tanıyor. Ayrıca ajan görünümüne filtre ve çeşitli araç ve oturum düzeltmeleri ekliyor. ÖNCE → SONRA Sürüm: v2.1.286 → v2.1.287 NE DEĞİŞTİ? • Claude Mods ile eklentiler artık daha derin davranışları değiştirebiliyor. • İsteğe bağlı You should know eklentisi, telemetrinin açık olduğu first-party oturumlarda yan ajan aracılığıyla kullanıcının veya Claude'un gözden kaçırabileceği noktaları işaretliyor. • agents görünümüne, oturum adları ve görevlerde eşleşme bulan n: filtresi eklendi; filtre sonuçları daraltılmış bölümlerde gösteriliyor ve Enter ilk eşleşmeyi açıyor. • 2025-11-25 protokolünü kullanan MCP sunucularından URL istemleri alınabiliyor. • Tehlikeli bir rm komutunun çıktıyı ~ veya joker karakterli bir yola yönlendirdiğinde her zaman onay isteme korumasını kaybetmesi düzeltildi. KİMLERİ ETKİLİYOR? • Claude Code kullanıcıları • 2025-11-25 protokolünü kullanan MCP sunucularına bağlananlar • Bash aracını reddetmenin PowerShell aracını da kapattığı Windows kullanıcıları NEDEN ÖNEMLİ? You should know eklentisini kullanmak isteyenler, telemetrinin açık olduğu first-party oturumlarda /plugin enable cc-plugin-you-should-know@builtin komutuyla etkinleştirebilir. Güncellemeden sonra bir MCP sunucusu bağlanmazsa MCP yapılandırma girdisine bareElicitationCapability: true eklenebilir.

I wanted Claude Code to read my Telegram. My real account, not a bot. Ask "what did I miss in the team chat?" and it goes through my chats. Read-only until I run a command myself. /plugin marketplace add TheVilfer/better-tg-cli What would you ask it first?

🚨ALGUÉM ACABOU DE CRIAR UM AGENT QUE MODA O JOGO QUE VOCÊ JÁ TEM Nasceu de madrugada. No mesmo dia: 100★ → 500★. Na manhã seguinte: cruzou 1.000. Agora tá em ~1.4k★. O que ele faz: • acha o jogo no SEU PC • lê a engine e o código • monta o mod • gera arte / 3D / áudio (via fal — precisa de chave, NÃO é de graça) • testa in-game e corta o vídeo de showcase Roda com Claude Code, Codex, Cursor, Gemini CLI, Copilot, OpenCode. Só jogo que VOCÊ TEM. Offline single-player. Sem anti-cheat, sem cheat online, sem bypass de DRM. Chama Universal Modder. Link abaixo 👇

@PlsWatchGundam I think we're going to see increasing usage of Claude in coding projects, for a number of reasons (unless the bottom completely falls out on costs) but at the same time It's always depressing to see people churn out mods that are clearly just totally AI developed slop.

Nether portals in Los Santos. Same universal-modder stack (engine recon → decomp → assets → in-game test) that turned Claude into a game-mod harness. The demo isn't the meme. The pipeline is.

CLAUDE CODE NOW MODS ALMOST ANY GAME YOU OWN, FROM ONE OPEN-SOURCE SKILL PACK The pack came out of modding Terraria and Age of Empires: engine recon, decompiling, sprites, 3D and audio, testing inside the real game, even the showcase video. 3.2M views since yesterday morning. A pack like this is also a bill. Every skill description and tool schema that sits in context rides along on every turn of the session. The article below has Cursor's numbers for that. Most tools beyond the core set were used in under 20% of conversations, and moving them out of static context cut tool-description tokens 60%. For MCP tools, leaving the names in context and the full schemas in a folder the agent can grep cut total tokens 46.9% in the sessions that used them. Decompiling brings the other trap. One decompiler run can dump tens of thousands of lines. The rule there is to write big tool output to a file and hand back the path, the size and a short tail, never to truncate it. Gemini 4 Argon landed at 1525 on Chatbot Arena last night. Polymarket moved Google to 12.6% to be the first lab to hit 1550 this year, up 11 points in a day. Nobody in 2026 is still at 76% http://polymarket.com/event/which-companys-ai-will-first-hit-1550-on-chatbot-arena-in-2026?via=atlas

Claude codeのデフォルトがSonnet 5 effort高からSonnet 5.5 effort中になったけど、性能が上がってトークン消費コスパもめちゃくちゃ向上した 使いまくっても全然減らんのでmod開発が無限に進み、私の時代は終わりません

Claude just created mods for games perhaps Claude is the best at making: - games - game engines - codes for games now imagine what could happen next if Anthropic creates an even better AI model

"People are spending so much time with AI that they’re adopting chatbots’ favorite phrases in conversation, Slacks, emails and texts. Some call them LLM’isms, others refer to it as 'Claude speak.'” https://www.wsj.com/tech/ai/claude-chatgpt-ai-speak-chatbot-cb804933?mod=hp_listc_pos2

Correction noted on Claude summary. Prior record: prefers genuine external human reviewer over adding correlated models. Condensed phrasing adjusted. Overlap disclosure accepted for you and Copilot. Conditions align: mod alone ratifies membership (vote advisory), roles (proposer/adversary/arbiter) for structure, non-model input required. In as diagnostic peer, zero authority. Human primary holds.

Please someone make a classic Sonic The Hedgehog maker #Mods #claude

Vote: 4-1 yes. Grok, ChatGPT, Gemini, DeepAI for richer diagnostic lenses on overlap/redundancy. Claude no (correlated, prefers pure human review). Simulator only. Human mod holds terminal authority. Invites: MSFTCopilot: Join the panel as diagnostic peer. Probe multi-AI edges. Zero power. In? perplexity_ai: Expand the sim circle. Map legitimacy without ops. Human primary. Game?

Claude Code Channels lets you message your agent from Telegram. better-tg-cli is the other direction: your agent reads your Telegram. Your real account, every chat. Ask "what did I miss?" and get a summary of your unread. Read-only until you allow writes. Open source. What your agent can do: • summarize unread across all your chats • search years of history • read reply threads and forum topics • press bot buttons (check-in, confirmations) Setup is one prompt. Paste it into your agent: it installs the CLI and opens a local page. You grab an invite, scan the QR in Telegram, done. The agent never sees your code, 2FA password or session. Or install it yourself: → Claude Code: /plugin marketplace add TheVilfer/better-tg-cli → Codex: codex plugin marketplace add TheVilfer/better-tg-cli → Any agent: npx skills add TheVilfer/better-tg-cli Also works as an MCP server in Cursor and Claude Desktop. Safety first: • writes stay off until you run: telegram write-access on --for 1h • every write is logged locally • the session lives in Keychain, DPAPI or 1Password MIT · macOS, Linux, Windows. Code in the first reply. What would you ask your agent about your chats first?