Instruments of the practice — not standalone consumer apps. Each tool puts one of the pillars into everyday use: AI governance and critical-thinking checks, structured prompting, by-the-book Scrum, and AI applied to consumer protection. All local-first, free while in beta, and built in-house.
Scans every prompt you type into Claude in real time — with ChatGPT and Gemini support in progress. It watches for sensitive data across 9 built-in categories — ID numbers, banking details, credentials, personal data, and more — plus any keywords you add, and stops you before you hit send. Locally. Privately. Your prompt text never leaves your machine. Use Alt+Shift+S to trigger a manual scan at any time.
All detection runs on your machine. No servers, no accounts, no prompt text stored — only a local, metadata-only audit log you can clear or export.
Catches sensitive data as you type — before the send button is ever clicked. Zero latency interception.
Designed with South Africa's Protection of Personal Information Act in mind. Tags each detection with the relevant section — it helps you spot regulated data, it doesn't replace your Information Officer.
Live on Claude now, with ChatGPT and Gemini support in progress — one extension across your AI tools.
Intercepts prompts on ChatGPT, Claude, and Gemini before sending. Choose from 12 frameworks across 4 categories — Simple (RTF, RACE, PACT, TAG), Scenario (COAST, TRACE), Complex (RISEN, COSTAR, CREATE), and Goal + Technique (SMART, CoT, Few-Shot). Auto-classifies your prompt, suggests the best framework, shows a live preview, and runs a quality checklist. Privacy-first: zero network requests, local storage only.
RTF, RACE, COAST, CoT, and more — covering simple, scenario, complex, and goal-based prompting styles.
Detects your intent and suggests the best framework automatically — no manual selection needed.
See the enhanced prompt before you send, with a quality score and actionable improvement tips.
Save, download, and revisit every prompt — all stored locally. No accounts, no cloud, no tracking.
Detects and highlights manipulative UI patterns in real time as you browse. Catches all 11 pattern types — sneaky sign-up checkboxes, cookie tricks, guilt-trip buttons, fake urgency, forced sign-ups, hidden catches, and more. Simple mode shows a plain-English health score with "What to do" tips. Expert mode gives the full technical detail.
Catches sneaky sign-ups, cookie tricks, guilt-trip buttons, fake urgency, hidden catches, and more.
Plain-English health score with actionable tips, or full technical detail for power users.
Instant local detection for known patterns, plus an optional Claude Sonnet 5 layer to catch the subtle ones.
Ignore findings you don't care about — per visit or permanently. You stay in control.
A self-contained, browser-based toolkit for running Scrum by the book — grounded in the 2020 Scrum Guide. Sprint hub with burn-down and velocity, planning, daily scrum, review, retrospective, planning poker, and guide-accurate timeboxes, all in one page. Everything is saved locally; no accounts, no backend.
Sprint hub, planning, daily scrum, review, retro and planning poker — no tab-hopping between tools.
Timeboxes per ceremony, scaled to sprint length, with an audio alert when time's up.
All state lives in your browser's local storage — burn-down, velocity history and retro notes survive a refresh.
Optional: turn a UI screenshot into user stories with Claude Sonnet 5, using your own Anthropic API key — which stays in your browser.
The JTC Tech portfolio is actively growing. New privacy-first tools and AI extensions are being built and will be added to this page as they launch. Check back regularly or get in touch to follow along.