Open source

Built for ourselves. Released for everyone.

We build the tools our own operations need — and open-source them when they're useful beyond us. These are the ones we use day-to-day.

TokenMaxx dashboard with quota windows, 7-day trend, and top projectsTokenMaxx usage view — token timeline and raw event log filtered by user, model, projectTokenMaxx projects view — AI spend grouped by project tree with token share and costTokenMaxx raw events table — full event log with model, project, tokens, cost per rowTokenMaxx models view — usage and cost broken down by provider and modelTokenMaxx subscriptions view — Claude Max and ChatGPT Pro plan utilisation and cost efficiencyTokenMaxx quota view — headroom across active quota windows for Claude Max and Codex ProTokenMaxx Wrapped — year-to-date AI usage summary with month-by-month token volume and provider splitTokenMaxx users view — workspace members and service accounts with 30-day usage

Developer tooling

TokenMaxx

v1.0.0 — Live

Track what your AI subscriptions actually cost — by project, by model, by day.

Captures token usage from Claude Code and Codex CLI, stores it in your own Supabase database, and presents a dashboard that breaks down AI spend by project, model, and day. No SaaS account required — your data stays in your own Supabase project. Live at tokenmaxx.structlabs.io with Supabase auth, live usage data, subscription tracking, and automated quota capture via browser session.

Next.js 15SupabaseCloudflare PagesTailwind CSSRechartsTypeScript
PSM dashboard with monitored platforms grouped by tierPSM incidents table with severity, status, and updated timestampsPSM routing rules — severity filters mapped to delivery venuesPSM platforms catalog with ingestion mode, health, services, and zonesPSM venues configuration — webapp, Telegram, and Slack delivery targetsPSM agent setup screen with config validation and agent prompt

Infrastructure

Platform Status Monitor

v1.0.0 — Live

Agent-configurable platform status monitoring for teams, agencies, and solo operators.

Converts SaaS and infrastructure provider status data into a concise operational dashboard. A Cloudflare Worker ingests and normalises incident feeds; Workers KV holds runtime state; a read-only Next.js dashboard surfaces it. JSON-first config stays readable, reviewable, and agent-editable. Telegram routing planned post-dashboard.

Next.jsCloudflare WorkersWorkers KVTypeScriptTurborepo

Need something built for your stack?

These tools started as internal needs. If you're dealing with a gap your current tooling doesn't cover, we build those too.

Contact usSee our services →
Built on