惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

freeCodeCamp Programming Tutorials: Python, JavaScript, Git & More
P
Proofpoint News Feed
宝玉的分享
宝玉的分享
人人都是产品经理
人人都是产品经理
博客园_首页
爱范儿
爱范儿
博客园 - 叶小钗
aimingoo的专栏
aimingoo的专栏
S
SegmentFault 最新的问题
MyScale Blog
MyScale Blog
阮一峰的网络日志
阮一峰的网络日志
IT之家
IT之家
Microsoft Security Blog
Microsoft Security Blog
Blog — PlanetScale
Blog — PlanetScale
博客园 - 【当耐特】
Y
Y Combinator Blog
量子位
博客园 - 三生石上(FineUI控件)
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
T
The Blog of Author Tim Ferriss
月光博客
月光博客
有赞技术团队
有赞技术团队
Apple Machine Learning Research
Apple Machine Learning Research
A
About on SuperTechFans

Hacker News: Show HN

PurrrrrFocus: Pomodoro Timer App - App Store Workflow Engine — Multi-Step Orchestration for Bun RapidPhoto: Pro Photo Editor App - App Store GitHub - DheerG/swarms: Achieve extraordinary results with claude code across a variety of tasks SPICE simulation → oscilloscope → verification with Claude Code — Lucas Gerads Show HN: VCoding – A 5 MB native Windows IDE with no dynamic dependencies Show HN: LLMs don't hallucinate because they're bad at math, it's the format GitHub - Agent-FM/agentfm-core: AgentFM is a peer-to-peer network that turns everyday computers into a decentralized AI supercomputer. AgentFM lets you run massive AI workloads directly across a global mesh of idle CPUs and GPUs. Show HN: Tracking Top US Science Olympiad Alumni over Last 25 Years GitHub - Potarix/agent-hub: One place to talk to all your agents Show HN: Runtime security for AI agents(injection,tool abuse, data exfiltration) GitHub - dubeyKartikay/lazyspotify: Terminal Spotify client for macOS and Linux GitHub - the-banana-tool/king-louie: Easy to use GUI Personal AI Assistant. Win/Linux/Mac. Show HN I made my vacation rental bookable by AI agents–no Airbnb, 0% commission GitHub - basteez/jsf-autoreload: maven plugin to enable hot reload on jsf projects uvm32/hosts/host-gdbstub at main · ringtailsoftware/uvm32 GitHub - labsai/EDDI: Config-driven engine that turns JSON into production-grade AI agents. Multi-agent orchestration, 12+ LLM providers, MCP/A2A protocols, RAG, persistent memory, and enterprise compliance (EU AI Act, GDPR, HIPAA). Built on Quarkus. GitHub - glitchnsec/fortyone-oss: AI Executive Assistant Platform Quickstart | Alien GitHub - muxshed/shed: One stream in, or many. Every destination, simultaneously. No cloud middleman, no per-channel fees, no limits. GitHub - ocrbase-hq/ocrbase: 📄 PDF/IMG ->.MD/JSON Document OCR API for PaddleOCR and GLMOCR. Self-hostable. GitHub - impactjo/home-memory: MCP server that lets your AI assistant remember everything about your home. GitHub - Sets88/dbcls: DbCls is a powerful terminal database client that supports various databases GitHub - neptun2000/heor-agent-mcp GitHub - SeanFDZ/macmind: Single-layer transformer in HyperTalk for the classic Macintosh RollQuation: Math Puzzles - Apps on Google Play GitHub - dropbox/witchcraft Show HN: Agent-cache – Multi-tier LLM/tool/session caching for Valkey and Redis GitHub - opentalon/opentalon: OpenTalon is an open-source platform built from the ground up in Go as a robust alternative to OpenClaw LinkedIn™ 职位抓取工具 - Chrome 应用商店
GitHub - mupt-ai/relaymux: lightweight, tmux-backed meta-...
byhong03 · 2026-06-18 · via Hacker News: Show HN

relaymux is a lightweight local metaharness for coding agents. Telegram is the remote control; tmux is where the actual agent work happens.

The background service listens for Telegram messages, but it does not hide agent runs in a black box. When relaymux launches an agent, it opens a visible tmux window on your machine so you can attach, watch, interrupt, or debug the run like any normal terminal session.

Telegram is the main supported interface. iMessage/SMS support exists, but it is beta.

Requirements

You need Node.js 20+, npm, tmux, and a local agent CLI such as pi, codex, or claude. The installer uses curl and tar by default; it only needs git as a fallback.

Install

curl -fsSL https://raw.githubusercontent.com/mupt-ai/relaymux/main/install.sh | bash

The installer downloads a source tarball, builds relaymux locally, and writes the CLI shim into a writable directory already on your PATH when possible. You do not need to clone the repo.

Set up Telegram

Create a bot with BotFather, copy the token, then run:

relaymux setup --telegram --telegram-bot-token '<telegram-bot-token>'

When prompted, open your bot in Telegram and send /start. relaymux stores the token in ~/.relaymux/secrets/telegram-bot-token, discovers your chat id, writes ~/.relaymux/config.json, and starts the background service.

Check it:

Use it

Send your Telegram bot a message like:

Open an agent in ~/code/my-app and inspect the failing tests.

relaymux passes the message to your configured local orchestrator. If the orchestrator launches an agent, that agent runs in tmux; the final reply comes back through Telegram.

Manual notification test:

relaymux notify --from test --reply-mode telegram --message "hello from relaymux"

tmux is the workspace

relaymux uses one shared tmux session named agents by default. Each launched agent gets its own tmux window. This is the core idea: Telegram starts and receives updates from work, but tmux keeps that work visible and recoverable locally.

Attach any time:

Detach with Ctrl-b d. Closing Telegram does not stop the tmux run.

Launch an agent manually

relaymux launch \
  --repo ~/code/my-app \
  --agent pi \
  --name inspect-tests \
  --prompt "Inspect the failing tests and summarize what is broken."

Then attach with:

iMessage/SMS beta

iMessage/SMS depends on a working local imsg command. The minimal setup is:

relaymux will try to show recent chats. Pick one, then text that chat to use relaymux. If chat discovery does not work, pass the chat id directly:

relaymux setup --imsg --chat-id '<chat-id-or-phone-number>'

Troubleshooting

relaymux doctor
relaymux status-launch-agent
relaymux status --history

Logs live in ~/.relaymux/logs. Config lives at ~/.relaymux/config.json.