DigiControl
The mission-control interface for your AI agent swarm. Dispatch work, monitor progress, and resolve blockers — all from a single dashboard.
DigiControl sits between your GitHub repository and your agent fleet. It watches for new issues and PRs, decides which agent to deploy, enforces resource limits, and surfaces anything that needs attention — so your swarm stays productive around the clock without you babysitting it.
Dashboard Preview
Dashboard screenshot coming soon
Architecture diagram · UI walkthrough · Feature tour
What DigiControl Does
Every piece of the orchestration loop — from issue intake to merge — handled automatically.
Real-Time Agent Monitor
See every agent — Lucy, John, George, Ringo — and what issue or PR each one is actively working. Status updates stream live via server-sent events.
Issue & PR Dispatch
DigiControl polls your GitHub repository and automatically routes open issues to the right agent based on labels, skills, and current load.
Parallel Agent Limits
Set a hard cap on concurrent agents and a spawn delay between launches. Prevent runaway compute costs without sacrificing throughput.
Stalled Work Detection
Issues that agents completed but left open are surfaced automatically. Lucy re-triages them — closing, reassigning, or escalating as needed.
PR Merge Automation
Pull requests tagged ready-to-merge are detected and merged by Lucy without human intervention. Merge conflicts are caught before dispatch and re-routed to the originating agent.
Configuration Panel
Adjust max agents, spawn delay, poll interval, and per-agent settings (model, token budget, timeout, skills) live from the settings sidebar — no restarts required.
Architecture
DigiControl runs as a lightweight Node.js service alongside your agents inside an isolated Kubernetes namespace.
Architecture diagram coming soon
GitHub → Poller → Dispatcher → K8s Agents → PR / Issue close loop