laosb

laosb / agentc

Public

Coding agents, easily containerized.

agent container containerization isolation security
22
0
100% credibility
Found Apr 10, 2026 at 22 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Swift
AI Summary

agentc runs AI coding agents like Claude Code and GitHub Copilot CLI in secure isolated environments with persistent profiles and project-specific memory.

How It Works

1
🔍 Discover agentc

You hear about agentc, a safe way to run AI coding helpers that stay organized per project without messing up your computer.

2
📥 Install easily

Run one simple command to download and set it up on your Mac or Linux machine.

3
📁 Open your project

Navigate to the folder with your code where you want the AI helper to work.

4
🤖 Start your AI helper

Type agentc run to launch it, and it prepares a secure space just for this project.

5
Pick your helpers
🧠
Claude for deep thinking

Great for explaining and improving code.

🚀
Copilot for quick commands

Fast for generating snippets and fixes.

6
💬 Chat and create

Talk to your AI helper, give commands, or open a shell—it remembers your settings and stays isolated.

✅ Code magic happens

Your AI assists with coding tasks safely, keeping everything organized and persistent across sessions.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 22 to 22 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 agentc?

Agentc runs AI coding agents like Claude Code and GitHub Copilot CLI in isolated containers, providing persistent profiles for auth and memory while isolating workspaces per project. Built in Swift, it supports Apple Containerization on macOS 15+ or Docker on macOS/Linux, with a simple CLI: `agentc run` launches agents, `agentc sh` drops into a shell, and profiles keep state across sessions. It pulls modular configs from a GitHub repo for easy agent swapping, solving messy local setups for github coding agent experiments.

Why is it gaining traction?

Unlike agentcore aws or agentcraft, agentc needs no cloud or complex gateways—just curl install and run, with native observability via container logs and per-project exclusions for clean isolation. Devs love activating multiple agents (`agentc run -c claude,copilot`) for coding agents comparison, benchmarking against cursor or VS Code agents without auth hassles or memory leaks. Its agentcubes-like pluggability shines in coding agents leaderboard tests, delivering agentcore memory and identity locally.

Who should use this?

AI-curious backend devs testing coding agents claude on macOS, Swift teams needing agentcore evaluations without Docker sprawl, or open-source contributors ranking github ai tools. Perfect for frontend folks dodging agentcore gateway setup, or anyone prototyping coding agents benchmark runs in secure sandboxes.

Verdict

Try agentc if you're into local coding agents ranking—its CLI and profiles make agentcon-style isolation dead simple, with strong tests covering Docker and Apple runtimes. At 22 stars and 1.0% credibility, it's early but MIT-licensed and mature enough for daily drives; watch for more agentcore observability integrations.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.