olavopitonjunior

AI Control Center — AI skill for Claude Code

AI community

Local-first, touch-first AI observability appliance.

How to install AI Control Center

This entry records only its repository, not the path inside it, so there is no exact command to give. Open olavopitonjunior/ai-control-center and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What AI Control Center does

Local-first, touch-first AI observability appliance. A Surface Pro control plane monitors Claude Code usage, system telemetry, AI coding sessions, and local + cloud automations across machines via a normalized agent API (Tauri + React + Fastify).

Alternatives in AI

Claudexor — Multi-harness control plane for Claude Code, Codex, Cursor, and OpenCode: quota-aware rotation across multiple 425 ★
Oc Claw — A desktop pet that monitors your AI coding agents (OpenClaw, Claude Code, Codex, Cursor, Gemini CLI, opencode 352 ★
Prismor — Self-hosted runtime control plane for AI agents 334 ★

README

AI Control Center

[![CI](https://github.com/olavopitonjunior/ai-control-center/actions/workflows/ci.yml/badge.svg)](https://github.com/olavopitonjunior/ai-control-center/actions/workflows/ci.yml)

A **local-first, touch-first AI observability appliance**. A Microsoft Surface Pro acts as a dedicated **control plane** that continuously displays and analyzes AI usage across your machines — provider quotas and limits, tokens and cost, active/historical AI coding sessions, system telemetry, and scheduled automations — for a **Mac (over Wi‑Fi)**, a **PC (over Ethernet)**, and more.

**Status: Milestones 0–5 complete.** A functional, local-first appliance. The agent collects **live** Claude Code usage (ccusage), system telemetry (Glances), local scheduled tasks (Windows Task Scheduler / macOS cron + launchd), and cloud automations (GitHub Actions / n8n / Supabase / Vercel) behind one authenticated, optionally-TLS API, and serves daily/weekly/monthly usage series. It advertises itself on the LAN via mDNS. The Surface discovers/registers machines, polls them, persists + downsamples history to SQLite, and shows real data with charts across Overview / Limits / Sessions / System / Usage / Automations / Insights — token/cost breakdowns, CSV/JSON export, and (with an optional quota ceiling) an **ESTIMATED** exhaustion forecast. Source provenance everywhere; no fabricated values.

**Real-world confirmations still pending** (built + unit-tested, not yet run on the target hardware/accounts in the dev environment): the **macOS agent** on a Mac, and the **cloud adapters** against live GitHub/n8n/Supabase/Vercel accounts. See `docs/BUILD_LOG.md`.

Architecture

 Monitored machine                          Surface Pro (control plane)
 ┌───────────────────────────┐              ┌──────────────────────────────┐
 │ ccusage (CLI)             ─┐             │ AI CONTROL CENTER (Tauri+React)│
 │ Glances (REST, localhost) ─┤  HTTP/JSON  │  Overview · Session