Skip to content

kagenti/humr

Repository files navigation

Humr logo

Run your own background agents on Kubernetes.
Isolated by default. Credentialed. Always-on.

Keep your coding agents running when you close the lid. Ship them to your team. Sell them to your customers. Humr gives Claude Code, Codex, Gemini CLI, or pi.dev an isolated Kubernetes pod, a credential-injecting proxy, a scheduler, and a Slack channel.

What you get

  • Zero-trust isolation — Every agent runs in its own pod with its own filesystem, network, and credentials. Outbound traffic routes through a proxy that injects real API keys; the agent never sees them. Network policy drops everything else. A compromised agent has nothing to steal and nowhere to go.

  • Always-on scheduling — Cron lives on the platform, not your laptop. Scheduled tasks look identical to human messages from the agent's perspective. Workspace and conversation history persist across restarts.

  • Slack-native channels — One Slack app, unlimited agents. Per-thread routing, identity linking via /humr login, per-instance access control. Your agents live where your team already works.

  • Bring your own agent — Claude Code and pi.dev ship as built-in templates. Codex, Gemini CLI, or anything that speaks ACP works too. No lock-in to one vendor's SDK or cloud.

Humr UI — agent chatting with Google Drive access

Guided Tour

git clone https://github.com/kagenti/humr && cd humr

Open your favorite AI coding agent in the repo and try:

Walk me through how Humr works step by step. I want to do a demo for myself.
Explain how things work on the way. Help me connect a model provider, create
an instance, add a connection to GitHub, and chat with an agent.

Once you're comfortable, go deeper:

Now show me the advanced stuff. Set up a Slack channel integration, create a
scheduled job, build a long-living agent with a heartbeat, and wire up an
MCP server.

Your agent has full context of the codebase, architecture decisions, and cluster commands.

Quick Start

Prerequisites: mise, a Docker-compatible runtime (Docker Desktop, Rancher Desktop, etc.), macOS or Linux.

mise install                # install toolchain + deps
mise run cluster:install    # create local k3s cluster + deploy Humr

Open humr.localhost:4444 (login: dev / dev), create an instance from a template, and start chatting. See the guide for cluster commands, credential setup, and Slack integration.

Learn more

  • Guide — credential setup, Slack integration, development workflow, architecture overview
  • Why Humr exists — the three problems every agent hits in production, how Humr solves each, and a 5-minute walkthrough

About

Run your own background coding agents on Kubernetes - isolated pods, injected credentials, scheduled triggers, Slack-native.

Resources

Stars

Watchers

Forks

Packages

 
 
 

Contributors