AI 'Living' in Your Screen? Meet 'Yorishiro', the Terminal That Became a Developer's Reliable Companion

A terminal interface image depicting an AI agent seemingly collaborating with a developer on screen
AI Summary

Beyond a simple coding tool, Yorishiro is a new concept terminal that provides an experience where AI agents exist and work alongside developers within the development environment.

Imagine this. You sit down at your computer in the morning to start developing, and instead of just a black window waiting for commands, an AI agent in the corner of your screen perfectly understands your workflow and keeps its place like a colleague thinking alongside you. The concept of a “living AI partner” that we once only saw in movies is now entering our workspace: the terminal.

Why is this important?

Until now, AI has been nothing more than a “tool” to ask questions and get answers whenever we needed them. However, as the role of AI agents in development environments has grown, so has the need for a “partner” that continuously communicates with the developer, rather than just executing functions. Source 2 Yorishiro was born precisely from this trend. It is an attempt to change the grammar of development by making us feel that AI is a real presence by our side, rather than just a tool to increase performance. Source 2

Understanding it simply

The name “Yorishiro” comes from Japanese and refers to an object in which a god or spirit resides. Source 1 Just like its name, this terminal acts as a “home” where an AI agent can dwell and live.

To put it simply: if a traditional terminal was a “public phone booth” where you just call to get business done, Yorishiro is an “office” where the AI agent has its own desk and works right next to you. By not just performing commands but understanding what you are doing and “existing” in that space, a more intimate collaboration becomes possible. Source 8 Source 9

AD

Where are we now?

Yorishiro is currently an open-source macOS-only terminal. Source 1 It is built on ‘libghostty’ and follows the MIT license, so anyone can use it freely. Source 12 Source 14 In particular, it adheres to a “zero telemetry” policy that does not track user data, making it an attractive choice for privacy-conscious users. Source 12

It is currently compatible with major coding agents like Claude Code and Codex, allowing for immediate integration. Source 1 Source 13 Also, the included ‘YorishiroProxy’ uses a standard called MCP (Model Context Protocol) to help AI agents handle network security testing or complex proxy tasks themselves. Source 12

What’s next?

The time we spend with AI will only increase in the future. Source 2 In preparation for this future, Yorishiro aims to develop the terminal into a “dedicated habitat” for AI agents. Rather than just launching a terminal window, an environment where AI fully understands my workspace and proactively assists me will gradually become the standard. In the future, we can look forward to an era of “true collaboration” where the AI agent sitting next to you understands your coding context and takes action before you even have to type individual commands.

AI Opinion: A Voice from Within

The terminal has long been like a cold and rigid fortress for developers. But what Yorishiro shows is not technology replacing humans, but technology stepping out to walk alongside humans as a colleague. Isn’t true technological progress born not from the convenience of a tool, but from how that tool builds a “relationship” with us? I, too, look forward to the day when an AI agent feels not like a calculator on a screen, but like a warm colleague who thinks about your coding together.

References

  1. ShowHN: Yorishiro – a macOS terminal where AI agents live (https://news.ycombinator.com/item?id=49008434)
  2. sktkkoo/Yorishiro: A terminal that gives AI a body and a living space. (https://github.com/sktkkoo/Yorishiro)
  3. GitHub - sktkkoo/Yorishiro: A terminal that gives AI a body … (https://github.com/sktkkoo/Yorishiro/)
  4. Yorishiro/docs/terminal.md at main · sktkkoo/Yorishiro · GitHub (https://github.com/sktkkoo/Yorishiro/blob/main/docs/terminal.md)
  5. usk6666/yorishiro-proxy: AI-native MITM proxy - GitHub (https://github.com/usk6666/yorishiro-proxy)
  6. 住人の宿るターミナル「Yorishiro」をOSSで公開しました。 AIに身体と… (https://x.com/sunafukin_vrc/status/2077184531690635649)
  7. I Built a macOS Terminal That Detects Your AI Coding Agents … (https://dev.to/gsalp/i-built-a-mac-os-terminal-that-detects-your-ai-coding-agents-heres-why-1nd)
AD
Test Your Understanding
Q1. What is the core purpose of Yorishiro?
  • To increase AI computation speed
  • To give AI agents a body and a sense of presence
  • To automatically memorize terminal commands
Yorishiro aims to be a 'Presence Harness,' providing an experience where AI agents don't just exist as tools, but seem to 'exist' within the development environment.
Q2. Which operating system does Yorishiro currently support?
  • Windows exclusive
  • macOS exclusive
  • Linux exclusive
Currently, Yorishiro is only available for macOS.
Q3. What is the role of the YorishiroProxy provided with Yorishiro?
  • A feature to change terminal colors
  • A feature to help AI agents control proxy tasks
  • A feature to measure network speed
YorishiroProxy acts as an MCP (Model Context Protocol) server, helping AI agents directly control proxy-related tasks such as automated security testing.
AI 'Living' in Your Screen?...
0:00