Skip to main content

Active projects

These are the six projects I am actively developing. Each one solves a real problem in my life and is shared so others can use or learn from them. They are local-first, run on infrastructure I control, and exist because I needed them.

For the full application collection including Fitness Coach and Blaster, see the application collection.

Muppitify Transcript Prep​

A free local app for Apple Silicon Macs and Windows that prepares coaching transcripts for AI, supervision, or mentor coaching, with ICF-aligned privacy by default. Transcript content is processed locally and never uploaded. No account, no sign-in, no analytics, no crash reporting.

What it currently provides:

  • Automatic detection. The detector scans and highlights every candidate for redaction. Names, emails, phone numbers, addresses, company names, and any other identifiers are replaced with stable placeholders so context survives and identity does not.
  • Full transcript editor. Search, manual redaction, text trimming, and ten steps of undo and redo.
  • Split a long recording. Break a transcript into two files by picking a split line, so each half can be redacted separately or handed to different reviewers.
  • Save and share. Export the cleaned transcript to a new file; the original is never overwritten.
  • Auto-updates. Signed and notarised on macOS, signed on Windows with an Extended Validation code-signing certificate, and installed with your consent only.

See the complete feature list.

The transcript redaction engine is published as a standalone open-source npm package so any project can reuse it. It provides deterministic PII detection and redaction for transcripts and text files, running entirely locally with no AI, no network access, and no cloud services. Name detection uses a 50,000-plus name dictionary with local grammar-based verification via compromise.js for context-aware precision, and no machine-learning models.

Published on npm: @muppit/shield-engine

Muppitify Transcript Prep dashboard with a drop zone for transcript files.

Muppitify Coaching Platform​

Deep AI analysis for professional coaches pursuing ICF certification. Session tracking, recording management, and progress toward ACC, PCC, and MCC credentials.

Available in four editions:

  • Desktop Edition (macOS app)
  • Server Edition (Requires Docker, runs on Windows, Linux, or QNAP NAS)
  • Vault Edition (macOS app, data stored as markdown in your Obsidian vault. Requires Obsidian to be installed)
  • K8s Edition (Kubernetes, runs in a container, separately licensed)

All core features work entirely offline after licence activation.

These foundation features work without any setup, internet, or external services:

  • ICF certification. Track hours, education, mentor coaching, evaluations, and exam results toward ACC, PCC, and MCC with a submission readiness predictor. Import and export ICF coaching logs. Manage peer coaching programmes. Compare certification levels side by side. Competency tagging and scoring against transcripts with talk time analytics
  • Session management. Dashboard, session logging in coach or client role, people and partnership management. Synced transcript playback with picture-in-picture. Import recordings from any video conferencing platform. Coach development areas with action plans
  • Privacy and portability. One-click privacy mode, full data export/import across editions, compliance log, and offline use after activation

AI analysis with a range of local and cloud providers for cross-session insights, competency assessment, and development seeding. Zoom recording import and Google Calendar sync. See the complete feature list.

Muppitify Coaching Platform overview.

School Platform​

A homeschool tracking platform for Western Australia with AI study unit generation, Google Calendar integration, and curriculum coverage tracking. Built with Next.js 16, PostgreSQL, and OpenAI, it replaces scattered spreadsheets and folders with a single self-hosted application.

What it currently provides:

  • Dashboard with stats, learning area grid, and upcoming dates
  • Activity tracking with multi-area tagging across 8 WA curriculum areas
  • AI study unit generation that turns seed ideas into cross-curricular learning adventures
  • Ask Bob Q&A assistant with conversation history and auto-categorisation
  • Google Calendar integration with privacy masking and activity linking
  • Evidence gallery for moderator meetings with photo upload
  • Mobile responsive design for iPhone and Android
  • Daily automated database backups with Slack notifications

School Platform dashboard overview.

Flow Control​

A self-hosted support tool for health and performance that predicts readiness and visualises flow state from real coding signals. It combines Cursor AI usage, GitLab pipelines, Apple Health data, Apple Music listening history, and exercise goals in one place.

What it currently provides:

  • Composite flow score engine computed from real coding signals
  • Sleep debt tracking and readiness prediction
  • Music-flow correlation analysis
  • AI health insights via local Ollama models
  • HealthSync iOS app for Apple Health data collection and companion Apple watch app
  • Slack alerts for work/break recommendations
  • Billing cycle management for Cursor usage

Flow Control dashboard overview.

MCP Task Server​

A Model Context Protocol server for AI-assisted task management with multi-agent coordination. Built to solve Cursor's memory limitations by sharing context between AI tools via a local file. What started as a workaround became a full task management system published as an npm package.

What it currently provides:

  • 44 tools for task management, dependencies, priorities, and progress tracking
  • Multi-agent coordination with Planner, Worker, and Judge roles
  • Shared context engine that bridges Cursor memories and MCP tools
  • Project scaffolding with memory bank templates
  • Slack integration for notifications and async conversations
  • Work activity statistics and flow state tracking

Published on npm: mcp-task-server

MCP Task Server tools overview.

Personal Data Hub​

A journaling system for lazy people. Books, shows, health, fitness, and more in a single queryable knowledge base built on Obsidian. Custom scripts, Chrome extensions, and a calendar plugin capture data automatically where possible and make manual capture fast where it cannot.

What it currently provides:

  • Book tracking with Google Books API, visual picker, AI summaries via Ollama
  • Show tracking with TMDB integration, episode tracking, and watch counts
  • Daily health metrics import (steps, HRV, sleep, weight) with incremental updates
  • Workout tracking with duration, calories, and workout type linking
  • Custom calendar plugin with month, week, and list views
  • Chrome extensions for Prime and Netflix watch history capture
  • Everything stored as plain markdown files with no cloud lock-in

Personal Data Hub book and show views.