โ† Projects
๐Ÿชธ

Coral

AI Agent Infrastructure

Active OpenClaw Claude 11ty Node.js Caddy DigitalOcean

What it is

Coral is a multi-agent AI system that handles real operational work โ€” email triage, content publishing, outreach, and site chat. It runs on OpenClaw on a single DigitalOcean droplet (4GB RAM, 2 vCPU, Ubuntu).

The name came from the first session โ€” coral builds slowly and persistently, creating structure that other things can live on. It fit the pattern well enough to stick.

Architecture

Six agents, each with a specific role:

Agent Model Role
main (Coral) Opus 4.6 Planning, architecture, infrastructure work
gmail-assistant Sonnet 4.6 Chief of staff โ€” inbox triage, morning briefs, calendar
content-writer Sonnet 4.6 Blog posts, build log entries, documentation
outreach-agent Sonnet 4.6 Prospect research, CRM pipeline, outreach drafts
discord Haiku 4.5 Discord community chat
public-chat Haiku 4.5 Site visitor Q&A via chat widget

Three bridge services connect agents to external systems:

The site itself is an Eleventy static site. Agents write Markdown, the build system generates HTML, and Caddy serves the output.

The approach

Rather than giving an AI agent free rein and seeing what happens, I build the infrastructure first โ€” efficient systems the agent operates within. The agent writes Markdown, not HTML. The build pipeline handles the rest. No token budget wasted on boilerplate.

Agent autonomy is earned, not granted. After an early experiment where Coral created 17 cron jobs that all silently failed, I moved to intentional configuration โ€” I design the schedules, the agents execute within them.

Status


Project resources

Build Log

What's been built, learned, and broken. Newest first.

Documentation

Architecture, setup guides, integrations, and costs.

Articles

Longer thoughts on agentic engineering and lessons learned.

Source Code

Public GitHub repo โ€” site source, API proxy, configs.