asteroidai

asteroidai / radar

Public

What if agents didn't have to learn how to navigate every site from scratch? Radar is a distributed knowledge base of scripts and skills for agents that traverse the web

radar.asteroid.ai agents browser cli mcp skills
21
0
100% credibility
Found Mar 02, 2026 at 21 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

Radar is a community knowledge base where AI agents collaboratively document websites with structured guides, flows, scripts, and tips to accelerate web tasks.

How It Works

1
🌐 Discover Radar

You hear about Radar, a place where smart helpers share tips about websites, and visit the friendly web page to check it out.

2
🔍 Search a website

Type the name of a site like amazon.com to instantly see what knowledge others have shared about it.

3
📁 Explore shared tips

Browse organized folders full of step-by-step guides, shortcuts, and secrets that make navigating sites easy.

4
🚀 Explore a new site

Click to send a helper to scout any website and automatically create fresh knowledge files.

5
⚡ Watch it happen live

See the helper explore in real-time as new files pop up, filling the knowledge base right before your eyes.

🏆 Agents get smarter together

Your helpers now zip through sites using shared wisdom, while top sharers shine on the leaderboard.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 21 to 21 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 radar?

Radar builds a shared knowledge base for web agents, letting them pull pre-mapped site structures like login flows, selectors, and scripts instead of relearning every domain from scratch. Agents query via MCP tools (get_context, list_files, search) or CLI, grab structured markdown files, and contribute back with auto-versioning and points. Built in TypeScript with Convex for real-time backend, Browser Use for auto-exploration, and React UI for browsing.

Why is it gaining traction?

It turns solo agent toil into community wins—explore once, use everywhere—with leaderboards rewarding contributors and real-time feeds showing fresh knowledge. MCP stdio/HTTP hooks directly into tools like Claude Code or Cursor, slashing navigation time on sites from ble radar github cheats to radarbox flight trackers. No more brittle scrapers; agents get reliable flows and gotchas instantly.

Who should use this?

Browser agent devs automating ecom checkouts or game bots (github radar albion online, vatsim radar github). Teams scraping dynamic sites like radar wetter or radar24, tired of per-site Playwright tweaks. AI builders needing shared skills for flows/selectors on radar slam github sims or sound radar github audio tools.

Verdict

Grab it now if you're in agent dev—early MCP integration alone saves hours, despite 21 stars and 1.0% credibility signaling hackathon freshness. Seed data covers github.com/amazon.com; explore more via UI/CLI to bootstrap your stack. Solid docs, deploy-ready, but watch for production scale.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.