Notes

Ollama now uses Apple's MLX framework under the hood on macOS, unlocking significantly faster performance on Apple Silicon. This accelerates local AI workflows from personal assistants to coding agents — a major win for developers running LLMs on M-series Macs without needing cloud APIs.

@lucaslovexoxo's Amore handles everything needed to self-publish macOS apps outside the App Store: Sparkle updates, code signing, notarization, DMG creation, and hosting. It includes a CLI for CI/CD pipelines and AI-assisted releases, S3-compatible storage, phased rollouts, and beta channels — removing days of manual Sparkle setup.

@helloitsaustin shares his setup for managing Google Ads with Claude Cowork — a custom plugin connecting to the Google Ads API via MCP, encoding common paid search workflows into skills. Works on desktop and Dispatch. He has moved 90% of his growth marketing work from chat to Cowork.

Cloudflare published a best practices guide for Durable Objects — especially useful for feeding to agents during early design stages and code review. Covers how to design around single-threaded, globally-unique instances with persistent storage.

@mynameistito built a terminal-native tool for creating Cloudflare API tokens — no more clicking through dashboard dropdowns. Select accounts, choose services, set read/write permissions, and generate tokens in seconds. Run it with npx directly from the command line.

@Jnatanh's pi-autoresearch-studio adds granular experiment-to-PR selection with auto-resolved dependencies to the Pi Coding Agent workflow. Built as a Pi extension, it lets you cherry-pick which experimental runs become pull requests instead of committing everything automatically.

@bcherny's tip for power users: Claude Code has deep git worktree support for running multiple agent sessions in parallel on the same repo. Use claude -w to start in a new worktree, or check the worktree option in Claude Desktop. Essential when you have dozens of agents working simultaneously.

@marcelpociot's Polyscope uses copy-on-write clones instead of complicated worktrees for managing multiple project checkouts. Way faster, uses less disk space, and includes all dependencies without complex setup scripts.

@_chenglou's new TypeScript library brings fast, accurate text measurement to the browser without touching the DOM. The pure userland algorithm can lay out entire web pages bypassing CSS and reflows — foundational work for the next generation of high-performance UI tooling.

@penberg from Turso clarifies their SQLite compatibility promise: databases are fully compatible with SQLite's file and WAL formats, you can always migrate back, and any incompatible features (like concurrent writes via BEGIN CONCURRENT) require explicit opt-in. When Turso falls short of compatibility, they treat it as a bug to fix rather than a feature difference to document.

@micLivs makes the case against vector search for most memory use cases. Their pi napkin tool uses BM25 and TF-IDF on plain Obsidian markdown files—no embeddings, no vector DB. On HotpotQA multi-hop questions it hit 99.8% recall and 79.6% F1 versus 90%/41% for the original and 29%/19% for Mem0. Sometimes a good map beats a complex pipeline.

Cloudflare Workers now lets you declare required secrets in wrangler.jsonc or TOML — Wrangler validates them locally, generates TypeScript types, and blocks deploys if any are missing. Small DX win that catches config drift before it hits production.

@M1Astra archived Anthropic's Claude Mythos blog post before it was taken down — a rare look at internal documentation about Claude's character, values, and the philosophical foundations behind how it was trained to engage with the world.

@appfactory's Agent-CI brings GitHub Actions workflows to local execution with a twist: when a step fails, it pauses and lets your AI agent fix the issue before retrying just that step. Uses the standard GH Actions Docker image but with near-zero caching latency, designed for the tight feedback loops agents need without burning through CI minutes.

Claude Code now supports .worktreeinclude files — drop one in your project root to automatically copy gitignored files like .env, .env.local, or config/secrets.json when Claude creates a worktree. Handy for keeping your agent's isolated branches functional without manual setup.

pi-lab lets you A/B test tools in the pi coding agent by intercepting calls, forking into isolated lanes, comparing outputs, and merging the winner back. Still experimental, but a promising approach for iterating on agent tooling without flying blind.

@aidenybai releases Expect — open-source CLI that lets agents test your code in a real browser. Point Claude Code or Codex at your app, get video recordings of every bug found, then fix and repeat until all tests pass. Works as a standalone CLI or agent skill.

@yazins's OpenOats is a macOS meeting companion that transcribes both sides of your calls in real-time and surfaces relevant talking points from your own notes. Runs entirely on-device with local speech recognition — no audio leaves your Mac. Pair it with Ollama for fully offline LLM suggestions, or connect to cloud models via OpenRouter. Recent updates added speaker diarization for up to 10 participants, slashed CPU usage from 84% to 17%, and introduced background mode with menu bar integration.

@connors introduces Sutro — an email client designed for people and their agents. Bring your own Claude, Codex, or OpenClaw agents and have them work with you directly in your inbox. Email is not going anywhere; it is just getting agentic.

After WIMP — @threepointone explores what happens when users can describe procedures in natural language that compiles to code. The historic split between programmers (who reshape the machine) and everyone else (who use pre-built interfaces) starts to dissolve. Every app becomes directly programmable on demand, not through menus but through conversation.