OtaskuHub
A central AI operations and portfolio-control dashboard for my personal AI ecosystem. It acts as the source of truth for app registry data, telemetry, agent activity, prompt governance, LLM usage, cost visibility, and human review workflows across projects such as OtaskuScout, OtaskuLang, and future Otasku apps.
OtaskuHub is designed like a miniature internal AI platform: apps report events, agent runs, agent steps, token/cost usage, failures, approvals, and operational health into one admin-facing command center.
- How to make AI activity inspectable instead of invisible
- How to connect prompt governance, approvals, and telemetry in one workflow
- How to expose useful operational state without giving tools direct database access
- How to keep personal AI infrastructure low-cost and serverless-friendly
- Supabase-backed authentication and admin protection
- App registry and API-key integration model
- Production telemetry endpoints
- Agent run and agent step tracking
- Metadata drawers for operational inspection
- Prompt-version visibility
- Pending approval review
- LLM cost summaries and app health checks
- Next.js and TypeScript application structure
- Supabase Auth with PostgreSQL/Neon or Supabase data patterns
- Vercel/serverless deployment posture
- OpenAI API usage reporting and cost visibility
- Telemetry events designed for operational dashboards
- Human-in-the-loop review surfaces for pending approvals
Leadership relevance
OtaskuHub connects implementation details to operating discipline: registry ownership, API-key boundaries, prompt versioning, cost visibility, failure tracking, and review workflows. It reflects the same delivery posture I use in platform work: make state visible, define ownership, and keep tradeoffs explainable.
Current status
Active / Building. Public descriptions should frame OtaskuHub as a personal AI operations and platform-control project, not an employer system or commercial SaaS product.