HN Showcase
Back to browse
GitHub Repository

Deploy and manage a fleet of OpenClaw AI assistants anywhere. Supporting hobbyist, team, and enterprise use cases.

17 starsGo

CongaLine – Self-hosted isolated AI agent fleet (OpenClaw, Hermes)

by zhendershot·Apr 8, 2026·4 points·0 comments
Visit ProjectView on HN

AI Analysis

●●SolidBig BrainSolve My Problem

Per-agent container isolation with separate networks beats shared-instance chaos.

Strengths
  • •Provider × runtime orthogonal design enables flexible deployments
  • •AWS SSM zero-ingress means no inbound ports required
  • •Promotion pipeline enforces same config across local, remote, production
Weaknesses
  • •Self-hosted AI agent management is becoming a crowded category
  • •Remote and AWS deployments default to OpenClaw only today
Category
AI/ML
Target Audience

Teams deploying multiple AI agents, security-conscious developers

Similar To

OpenClaw · Hermes Agent · LangChain Serve

Post Description

We built CongaLine to solve a specific problem: give teams and individuals self-hosted AI assistants without collapsing everyone into a shared instance. Each agent runs in its own Docker container with isolated networks, secrets, and config. Security is the primary design constraint.

Three deployment targets via a single Go CLI (conga):

- Local: Docker Desktop, no cloud required - Remote: any SSH-accessible Linux host (VPS, Raspberry Pi, bare metal) - AWS: zero-ingress via SSM, no inbound ports

Runtime is decoupled from deployment. OpenClaw (Node.js) and Hermes Agent (Python) ship out of the box - same infrastructure, your choice of agent. Multi-runtime is fully wired on local today; remote and AWS default to OpenClaw. More coming soon.

Channels are optional. Every agent gets a web UI via SSH/SSM tunnel. Slack and Telegram are both supported, each with a dedicated router container that fans events to per-agent containers. Two agent types: user agents (DM-based) and team agents (channel-based).

The name is literal: spiny lobsters travel in single-file conga lines during seasonal migration, in physical contact, reducing drag and sharing protection. A procession of isolated-but-coordinated agents.

Apache 2.0. Repo has Terraform for AWS, a Go CLI, and a Node.js Slack router.

Would love feedback on the provider abstraction. We're planning Kubernetes and GCP next.

Similar Projects

Developer Tools●Mid

ClawHQ – Fleet management dashboard and skill marketplace for AI agents

Real-time fleet view with heartbeat monitoring, drag-and-drop task kanban, and a unified agent chat paired to a one-click skill store and 80/20 creator split — that combo is useful and not something you see everywhere. Practical execution looks solid from the UI, but the product’s reach depends on OpenClaw adoption and how comfortable folks are exposing gateway URLs and installing third-party skills.

Bold BetSlickNiche Gem
zchmael
104mo ago