This guy building an AI-native macOS dock with custom modes like writing and building to keep him focused. His agent can edit the modes on its own. THE DOCK IS BECOMING AN AGENT NOW.

Most people are using AI coding agents completely wrong Instead of letting Claude or Codex blindly write code, Superpowers makes them plan, test and actually think through what they’re building Get started in 3 steps: * Install: Add Superpowers to your coding agent * Plan: Tell it what you want to build * Execute: Let it write, test and debug your code step by step Less broken code. Less wasted time. Better results. Repo: github.com/obra/superpowers⁠

Claude has a feature that most people aren’t taking advantage of You can make it turn a huge, confusing codebase into something you actually understand You don’t need to be an experienced developer Try this prompt: “Act as a senior developer onboarding me to this codebase. Map out the project structure, explain how the main components work together, and identify the 5 most important files. Show me where to make common changes, flag potential issues, and create a simple roadmap for understanding the entire project.” Instead of spending hours figuring out how everything works, let Claude walk you through it Save this before opening your next unfamiliar GitHub repo

5 GITHUB REPOS FOR ONCHAIN DATA ANALYSIS: 1.⁠ ⁠Dune API github.com/duneanalytics/dun… 2.⁠ ⁠web3.py github.com/ethereum/web3.py 3.⁠ ⁠ethers.js github.com/ethers-io/ethers.… 4.⁠ ⁠Subsquid github.com/subsquid/squid-sd… 5.⁠ ⁠The Graph github.com/graphprotocol/gra… Onchain data is public. The edge comes from knowing how to actually use it

Lock in chat Memecoins are so back

MultiPost is a free browser extension that publishes your content to multiple social platforms in one click. >⁠Write once, post everywhere >⁠Works on Chrome and Edge >⁠Open source under Apache 2.0, already at 3.7K stars ONE POST. EVERY PLATFORM. NO SCHEDULER SUBSCRIPTION. ARE PAID SOCIAL TOOLS ABOUT TO GET REPLACED? github.com/leaperone/MultiPo…

Liquid AI just dropped LFM3, a family of 6 tiny models across text, audio, vision and omni. LFM3-VL-2B matches Gemma 4 12B, which is 6x larger, with just 1.9B parameters. LFM3-1.7B sits #1 on the Pareto frontier for on-device intelligence now.

That 40-video YouTube playlist you saved and never finished? Turn it into a course in one prompt. Open Claude Code in an empty folder and paste: "Install yt-dlp if it's missing. Download only the English subtitles, not the videos, for every video in this playlist: [link]. Then build me a course: 1.⁠ ⁠Split it into modules, in the right order, even if the playlist isn't. 2.⁠ ⁠For each module: a one-page summary, the 5 key ideas, and timestamps linking back to the exact moments in the videos. 3.⁠ ⁠Cut the filler: sponsor reads, repeated intros, off-topic tangents. 4.⁠ ⁠A 10-question quiz at the end of each module, with answers explained. 5.⁠ ⁠Save everything as a clean website I can open in my browser." Hours of video, done in an afternoon.

Your X account already knows which of your posts work. You just never read the data. Download your archive: Settings → Your account → Download an archive of your data. It can take a day to arrive. Unzip it, open Claude Code in that folder, and paste: "Read data/tweets.js. Spin up 5 subagents in parallel: 1.⁠ ⁠Topics: group my posts by subject and rank each group by average likes and reposts. 2.⁠ ⁠Hooks: compare the first lines of my top 20 posts with my bottom 20. What do the winners do that the losers don't? 3.⁠ ⁠Timing: which days and hours my posts do best. 4.⁠ ⁠Formats: threads vs single posts, lists vs stories, with links vs without, with media vs without. 5.⁠ ⁠Dead weight: the topics and formats I keep posting that never work. Merge it into one page: what to post more, what to stop, my 5 best hook patterns, and 10 new post ideas written in my winning style." You'll learn more in 10 minutes than in a year of guessing.

Power's out, internet's down, someone's burned. Your AI is useless without a connection. Survival LLM runs entirely on your own computer with no internet. Ask how to keep food safe in a blackout, treat a burn or purify water, and it answers from an offline library of medical, emergency and how-to articles, showing the exact passage each answer came from. You can add your own files too: your generator manual, your medicines list, your family's emergency plan. Set it up today, while you still have internet. github.com/anonnerdfrenzy/su…

Motion capture for a dragon used to need a studio. Now it costs $0.26. anymove turns one AI-generated video into a real animation for any rigged creature: dragons, giant spiders, wolves, fire demons, with wings, tails and eight legs included. An AI video model films the creature from two angles in one frame, a vision model finds every joint, and the two views pin each joint in 3D. The rig follows frame by frame and exports straight into any game engine. Under a minute per clip, no hand edits. Needs an NVIDIA GPU. github.com/patch-gg/anymove

You can now check your own phone for Pegasus, for free. Zero Breach Toolkit looks for traces of known government-grade spyware, including Pegasus, Predator, Graphite and Candiru, on Windows, Mac, Android phones and iPhone backups. It's read-only, makes zero network requests during a scan, and never uploads anything. Every report says how strong the evidence is and what it couldn't check. A clean result is never sold to you as "you're safe." No technical experience needed. The guide walks you through it in about 20 minutes. github.com/interplains/Zero-…

Your coding agents forget what the last one did. A new CLI called agentlog is built to fix that. It lets one agent search another's past session logs from Claude Code, Codex or GitHub Copilot CLI, by directory, session ID, time or task words, and returns a bounded preview with follow-up commands for edits, subagents or grep. Per the README: it never writes to or deletes logs, makes no network connections, and has no dependencies. Those are the author's claims; I haven't audited the code. Caveats: it was built mainly on macOS, Linux is untested on real logs, broad reads can cost tens of thousands of tokens, and logs can contain secrets, so output should be treated as data. Install is a curl-to-shell script (or go install); read it first. Apache-2.0, created Oct 27. github.com/Ryheff24/agentlog

Miw’s Town is a new repo that turns Claude Code sessions into 3D characters in a pastel town, sending them to the Library for file reading, the Writing Studio for editing and the Bug Clinic on failure, with finished sessions going home to rest, all reported through a hook script with no polling. → Hook reports 7 Claude Code events to a local server at 127.0.0.1 → Start headless tasks from the UI, customize characters and visit a cat room → Protections include localhost only, a per-run token and task folders limited to your home directory → No license shown, repo is hours old, read hook.js and server.js before adding the hook Repo: github.com/musictwo2117/MiwT…

Every MCP server you add to an AI coding agent is code and instructions your agent trusts. safemcp, a new MIT-licensed CLI, audits that setup. It reads configs for Claude Code, Cursor, Codex, VS Code, Windsurf and others, and flags auto-approval settings, unpinned npx/uvx packages and risky instruction files. With --live it also lists each server's tools and checks them for hidden instructions, invisible characters and tools that change after approval. Caveats: the repo is days old with few users. --live starts your configured stdio servers, which runs their code, and OAuth servers are skipped. 'Sends nothing anywhere' and the large tuning dataset are README claims I haven't verified. Try it on the included demo config first. github.com/0x1903/safemcp

Claude Opus 5.5 made a one-minute 3D animated film from a single prompt that read like a creative director’s brief, with hard constraints against text, templates and Remotion and a closing line that said “don’t create slop, I want art and creativity.” → Asked for a researched story, original 3D scenes and sound from free online assets instead of code-generated audio → The video and post couldn’t be verified independently so treat the one-prompt claim as unconfirmed → The prompt itself is worth studying regardless of whether the result was cherry-picked Reddit: reddit.com/r/ClaudeAI/commen…

AI agents editing your Word docs usually means a rewritten file and no way to see what changed. Quillwork, a new open-source tool (alpha, 0.0.x), makes the agent's edits show up as normal Word tracked changes, attributed to the agent, so you accept or reject each one. It splices edits into the .docx instead of regenerating it. Per its README, a no-op open and save is byte-identical on 19 of 19 test files, versus 0 of 19 for python-docx. That is the author's own benchmark and I haven't reproduced it. Caveats: only .docx/.docm, no footnotes, hyperlinks or image edits yet, a known bug with vertically merged table rows, and breaking changes expected. It ships an MCP server, a Python package and a Rust core, under Apache-2.0. github.com/gardan4/quillwork

GM CT ☕️

Drop a video into Claude Code, type /effects and the effect you want. That's the whole edit. /effects orbit from 3 to 6 seconds /effects spotlight when I say "kill the lights" /effects comment-box at the end, typing CLAUDE There are 45 effects: brackets that lock onto your face, a giant word dropping in behind you, a checklist that ticks off as you talk, a Follow button that flips to Following, and more. You can stack several in one go. It checks the frames first so nothing covers your face, and your audio stays untouched. Install: git clone github.com/Vibha-Ramprakash/… ~/.claude/skills/effects

You can now run Claude Code from your phone, by voice, while you're out walking. Say what you want done and on which computer. pitupitu starts the session there, keeps an eye on it, answers what it can, and pings you when it's done or stuck. Questions show up as buttons you can tap. Works with Claude Code and Codex, on any of your machines. Nothing opens a port on your computer. It dials out, and it never touches your Claude or Codex login. Install on each machine: curl -fsSL pitupitu.chat/install | sh pitupitu setup Then open pitupitu.chat/app on your phone and talk.