kju4q

5 repos the AI builder community can't stop talking about, and what to steal from each

These five kept coming up across the AI builder community this week, so here is what each one is and the one thing worth taking from it.

---

1. everything-claude-code

https://github.com/affaan-m/ECC

kju4q

5 things about Fable 5.1 that matter more than the benchmarks

Fable 5.1 launched in September 2026, and almost everyone covered the benchmarks. These are the five changes that actually affect how you build.

---

1. Your Claude Code sessions can message each other

• Shipped quietly in Claude Code 2.1.224: separate sessions on the same machine can communicate directly.
• Two tools do the work: ListAgents finds other active sessions, SendMessage delivers between them.

kju4q

3 High-Leverage Claude Techniques Most People Don't Use

1. Define a Research Subagent Instead of Prompting for One

Most people let Claude read multiple files directly in the main session. This quickly bloats the context window and reduces reasoning quality. Advanced users delegate research to subagents, but they don't do it with a text rule anymore. They define the subagent as a real, reusable agent.

Why it works: Subagents run in their own isolated context window with their own system prompt, tool permissions, and model. The main session only ever sees the final summary, never the intermediate file reads. Your primary context stays focused on decision-making while the research happens off to the side.

How to implement it: Run /agents inside Claude Code and create a project-scoped agent. It lives as a Markdown file in .claude/agents/ (or ~/.claude/agents/ for all projects), so it's version-controlled and shareable with your team:

kju4q

Nested Agents: the limits, the math, and the settings

Three things changed in Claude Code between July 1 and July 24, 2026. Agents can hire their own agents again, there is a hard ceiling on how many, and they all run in the background by default.

Everything below is quoted from the official changelog with version numbers, so you can verify any of it at code.claude.com/docs/en/changelog.

What actually changed

Nesting is back on, at depth 3 (v2.1.219, July 24)

kju4q

The Room

Three sessions, one idea, and none of them can see the other two.

Most ways of pressure testing an idea with AI happen in one conversation, so every answer is shaped by the answer before it. A room is different. You ask three separate sessions three opposing questions about the same idea, and none of them knows the others exist. It is the standard a journalist uses before printing something. Two independent sources who never spoke. If they land on the same thing anyway, that is not agreement, it is confirmation.

Before you start

Open three separate chats. Not three tabs in the same conversation, three actual chats, or they can see each other and the whole thing collapses.

kju4q

Three things already in your claude code that most people haven't touched. The setup for each, from my real workflow.

---

1. Routines

A routine is a scheduled cloud agent. It is not a local cron job. Each run spawns an isolated session in Anthropic's cloud with its own git checkout. It cannot see your laptop, your local files, or your local env vars. Everything it needs has to come from the repo it clones and from the prompt.

Here is the repo-health check I set up, step by step.

kju4q

Social Stats Dashboard

A local, single-file dashboard for tracking your Instagram and TikTok post stats. It's one index.html file. There is no server, no build step, and no account. Your data never leaves your machine — it lives in the file (or your browser's local storage), nowhere else.

Use it

1. Download index.html.

kju4q

Find What Your Prompt Is Missing

Three ways to close the gaps claude fills with guesses. From the video, made specific so you can run them today. Concept credit: a Claude Code team member's field guide on finding your unknowns.

---

1. The Blindspot Pass

kju4q

Run Claude's Most Powerful Model at Half the Price

The two official patterns from Anthropic, plus the exact setup I used on camera. Benchmarks are Anthropic's own: advisor pattern scored 92% of Fable 5's quality at 63% of the price (SWE-bench Pro), orchestrator scored 96% at 46% (BrowseComp).

The idea in one line: stop letting the expensive model do the typing. Put it where judgment matters, let cheap models do the volume.

---

kju4q

The Claude Code Playbook

How Boris Cherny, the creator of Claude Code, actually uses it day to day. Curated from his public thread and interviews, organized in one place.

All of this is his, sourced from his own posts and talks. Links at the bottom.

---

1. Run many Claudes at once