Sunday, 22 November 2026

The news in one minute. Three languages, one paper.

Nvidia down nearly 48% from peak as AI bubble fears spreadInkling 2 Mini: a layer-looping model that runs on a laptopYouTube Reportedly Set to Pay $3 Billion to Settle Addiction ClaimsAnthropic lists on Nasdaq, shares jump 33% above IPO priceJensen Huang Quits Nvidia Amid Chinese Spy SuspicionsFirecrawl bundles scraping, search and data for AI agentsSuperpowers: skills that make coding agents work with a processCLI-Anything: turning software into tools for AI agents
Open sourceGitHub Trending

Superpowers: skills that make coding agents work with a process

A Shell skills set gained over 2,000 stars in a single day

What made a skills set written in Shell gain more than two thousand stars in a single day? Superpowers by obra on GitHub turns a coding assistant into something that works with a process, instead of diving straight into writing code.

The project is built by Jesse Vincent and the Prime Radiant team, and described as a software development methodology for coding agents. When it sees you are building something, the agent stops to ask what you really want, drafts a spec, shows it to you in short chunks for review, then makes a plan. Once you agree, subagents handle each task and check their own work.

Three numbers

  1. 2.03K stars01
    Stars today

    Stars gained in one day on GitHub Trending

  2. 322.2K stars02
    Total stars

    Total stars of the obra/superpowers repository

  3. 2-5minutes03
    Per task

    Length of each small task in the implementation plan

Today the repository gained 2028 stars, bringing the total to 322222. Each task in the plan is cut small, about two to five minutes. According to the documentation, the agent can work autonomously for a couple of hours at a time without deviating from the plan. The skills install on many environments, such as Claude Code, Codex, Cursor or Gemini CLI.

This suggests users want agents with discipline, not just fast code. Testing first and a review after each task are set as mandatory workflows. The appeal may also come from the same skill set running across many agents. The question is: would you let an agent ask detailed questions before writing the first line of code?

Why it matters

  1. Process comes before coding speed
  2. Testing and review become mandatory steps
  3. One skill set runs on many agents

Would you let an agent ask questions before coding?

AI-generated illustration, not a photo of the event
AI-generated illustration, not a photo of the event
Figures automatically checked against the original story
  • ✓2.03 K sao: Sao hôm nay: 2028
  • ✓322.2 K sao: Tổng sao: 322222
  • ✓2-5 phút: Breaks work into bite-sized tasks (2-5 minutes each)

Story and video compiled by AI from the source above, with an AI voice. Figures are checked automatically against the original.

Keep reading