Michaelliv

Michaelliv / mercury

Public

🪽 Mercury — There are many claws, but this one is mine.

92
1
100% credibility
Found Mar 08, 2026 at 92 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

Mercury is an open-source personal AI assistant that connects to WhatsApp, Slack, and Discord to provide context-aware responses with persistent memory organized into user-defined spaces.

How It Works

1
📰 Discover Mercury

You hear about Mercury, a friendly AI sidekick that chats right in your WhatsApp, Discord, or Slack groups.

2
📦 Set up your assistant

Create a new folder for your assistant and run a simple setup command to prepare everything.

3
🧠 Connect an AI brain

Sign in with your favorite AI service so your assistant can think and respond smartly.

4
📱 Link your chat apps

Tell your assistant which WhatsApp, Discord, or Slack accounts to watch for messages.

5
📂 Create personal spaces

Make separate memory zones like 'work' or 'family' to keep conversations organized.

6
🔗 Link chats and say hello

Connect your group chats to spaces, mention your assistant, and watch it reply for the first time!

7
⚙️ Run it always-on

Start your assistant as a background helper that never sleeps.

🎉 Your AI lives where you chat

Now your personal assistant remembers everything, handles tasks, and chats seamlessly across all your apps.

Sign up to see the full architecture

6 more

Sign Up Free

Star Growth

See how this repo grew from 92 to 92 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is mercury?

Mercury deploys personal AI agents into your WhatsApp, Slack, and Discord chats, handling conversations across platforms via user-defined "spaces" that share memory and Obsidian-compatible vaults. Built in TypeScript with Docker isolation and the pi runtime, it processes messages, media like images and voice notes, and scheduled cron tasks through a simple CLI: install via npm, run `mercury init`, auth with providers like Anthropic, then `mercury run`. Tired of switching apps for AI? It lives where you chat already.

Why is it gaining traction?

Unlike standalone bots or bloated LLM wrappers, Mercury unifies chats into isolated spaces with RBAC permissions, KB distillation for chat knowledge extraction, and TypeScript extensions for custom skills or CLIs – all auto-baked into Docker images. Devs dig the zero-config multi-platform setup, media handling, and pi-powered agents that read/write workspaces directly. Freddie Mercury sang "there must be more to life than this" – this adds AI without the hassle.

Who should use this?

Solo devs managing family groups on WhatsApp, engineering teams bridging Slack/Discord workflows, or anyone building personal ops bots with cron reminders and file I/O. Ideal for tinkerers extending via `mercury add npm:package` or git repos, not enterprise-scale deployments.

Verdict

Promising early project (86 stars, 1.0% credibility) for chat AI experimentation – solid CLI/docs make it approachable, but low maturity means test thoroughly before production. Grab it if you want Mercury on GitHub without Cisco's enterprise baggage.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.