Screens

Train, recover and understand your own data, on the phone and on the wrist.

A training and recovery app for phone and Apple Watch: structured workout execution, sleep, readiness and recovery guidance, correlations across sleep, training, mood and weight, strength progression, AI blood-test extraction, coaching and a social layer, with a premium subscription.

Watch your AI agents work, and talk to them, in a 3D office.

A Gather-like 3D office for watching AI agents work: running agent sessions are the characters, each live session gets a room, subagents walk in as named engineers, and you walk around with WASD, read the room log and type instructions to the agents. Ships an MCP server so agents can join, speak, move and check their inbox.

Proofreads and rewrites your text in any Windows app, with a local model.

A writing assistant that floats over any Windows app: select text anywhere, press a shortcut and get a word-level diff of corrections; Replace pastes back over the live selection. Runs on a local model or ChatGPT, and builds a personal writing profile across 19 error categories with drills and a stats dashboard.

Read Markdown instantly and compare two documents side by side.

A fast Markdown reader that also diffs two documents: Markdown, JSON, PDF and DOCX side by side, with tabs, an edit mode with live preview, session restore and Explorer right-click integration. Built on Electron; cold start around 390 ms.

Keep every timezone you work with on screen, in one floating widget.

A floating, always-on-top world-clock widget: a tray control panel to add clocks, pick a base timezone, switch layouts, compare timezones and, optionally, show FX parity per clock.

Design and print labels on Niimbot thermal printers, from the browser or the phone.

A label editor for Niimbot thermal printers, in Portuguese: compose labels from text, QR codes, barcodes, icons and shapes in millimetres, preview the exact monochrome output and print over Bluetooth. The native app wraps the editor with a BLE bridge.

Compose screens from a product’s real components, so a mockup behaves like the product.

A drag-and-drop screen designer built for an internal design-system team. Components come from the product’s own Angular library and are instantiated live on a 12-column grid, so every input, dropdown and table in a mockup is the real thing with its real behaviour. An inspector edits each component’s declared inputs; screens preview at device widths and export to PNG, JSON or generated Angular code. Comments, templates, snapshot diffs and multi-screen flows sit on top of the canvas.

An OpenID Connect provider with passkey login and no passwords, shared by my apps.

A single sign-on service built from the spec: OpenID Connect with authorization code + PKCE, EdDSA-signed tokens, rotating refresh tokens with reuse detection, and WebAuthn passkeys as the only credential, backed by a one-time recovery code. Its first relying party is an admin editor for a Markdown content API: WYSIWYG editing, EN/PT bodies, live preview, and publishing that triggers a static-site rebuild through a deploy hook.

An Avatar fan game: Aang fights through the Spirit World, one run at a time.

An isometric action roguelite in Godot 4.7, set in the instant between Azula's lightning and Katara's healing water. Aang bends all four elements in hand-painted rooms, picks blessings from past Avatars that reshape each run, and faces bosses between visits to a hub. A free fan game: Avatar and its characters belong to Paramount and Nickelodeon, and nothing here is sold.

A lighting crew for your webcam: drag the lamps around your own face and room, mid-call.

Your call, lit like a studio. Up to four coloured lights fall on your real room, with a soft shadow on the wall, and the picture follows you as you move. A Windows app holds your setups, and a Chrome extension brings the same light into Google Meet and Zoom on the web. Open source.

Let your AI agent record a clean demo of one app, with nothing on top of it.

An MCP server that lets an AI agent record one app window: the video shows only that app, and windows dragged on top of it never reach the recording. It captures a whole window, a part you drag out or a monitor into a constant frame rate MP4, with audio off unless asked. A desktop app laid out like the Snipping Tool shares the same recorder, so you see a bar with an AI badge whenever the agent records and can stop it. Tested on Windows; the macOS backend is experimental.

Score what an AI coding agent delivered against its ticket, and refuse the PR when it falls short.

A quality gate for AI-written code: each delivery (a PR, a branch or a working tree, plus its ticket) is scored on weighted dimensions such as requirement coverage, completion, correctness, verification, scope, PR fidelity and conventions. Deterministic checks run first, LLM judges score what they cannot, and blockers fail a delivery regardless of score. It runs before the agent opens a PR, and again on every pull request through a GitHub webhook.