beam-cloud

beam-cloud / airstore

Public

The filesystem for AI agents

www.airstore.ai ai claude-code context filesystem fuse
158
8
100% credibility
Found Feb 05, 2026 at 25 stars 6x -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Go
AI Summary

Airstore turns data from apps like Gmail and GitHub into easy-to-browse folders that AI tools can read and analyze using everyday language.

How It Works

1
👋 Discover Airstore

You find Airstore online and sign up for a free account to bring your data to life for AI.

2
🔗 Link your everyday apps

Connect services like your email, GitHub, or notes so your real data is ready to use.

3
🗂️ Describe folders in plain words

Create special folders by typing simple requests like 'unread emails' or 'open issues' – magic happens!

4
💻 Add the helper app

Download a tiny app to your computer and sign in once to unlock everything.

5
📁 Open your data world

Mount the folders on your desktop – now your data looks like regular files you can browse.

6
🤖 Ask AI anything

Jump into the folder with your AI helper and chat: 'sum these invoices' or 'find unpaid ones'.

✨ Insights appear instantly

Your AI pulls smart answers from your real data, saving hours of manual work.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 25 to 158 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 airstore?

Airstore builds a FUSE filesystem for AI agents, pulling data from Gmail, GitHub, Linear, Notion, and Slack into mountable directories via natural language queries. Create smart folders like "open PRs needing review" or "invoices last week," which sync as readable files—`ls ~/airstore/gmail/invoices` shows `.eml` files ready for `cat`. Built in Go, it exposes MCP tools (filesystem, Wikipedia, GitHub) as CLI binaries in `/tools/`, chainable like `~/airstore/tools/wikipedia search "agents" | grep einstein`.

Why is it gaining traction?

Unlike API wrappers or vector stores, airstore lets agents use familiar file ops—Claude reads, searches, and processes data without custom prompts or SDKs. Scoped smart folders prevent over-sharing (e.g., inbox subset only), and team workspaces share contexts securely. Local mode runs self-hosted with FUSE, dodging cloud lock-in for agents filesystem context engineering.

Who should use this?

AI devs prototyping file system agents or filesystem for agents with Claude/Anthropic tools. Teams doing context engineering who want `claude` to summarize GitHub issues or Gmail threads directly. Claude Code users tired of manual data dumps, or github mcp filesystem experimenters.

Verdict

Early with 73 stars and 1.0% credibility score, but strong docs, demo video, and E2E tests signal promise for airstore meta workflows. Prototype if building agents filesystem; hold for production until more adoption.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.