Projects and experiments
A directory of browser AI experiments, web platform research, and tools that you can open and use.
Actively developed and hosted on GitHub Pages.
Autonomous SDLC Fleet
A schedulable fleet of software-development-lifecycle agents that run locally under your existing developer session auth (pi, claude, antigravity) and in CI on GitHub Actions. Deterministic scanners enumerate; AI models triage, synthesize, and judge.
Voice AI Coding
The voice-first coding environment. Speak naturally with an AI coding model that reads, reasons aloud, and modifies code in real time—with immediate in-browser file reflection, duplex audio, and ACP harness bridge.
Browser AI Extension
A privacy-first in-browser agent harness for Google Chrome. Seamlessly embeds autonomous coding agents, browser automation tools, and MCP servers directly into Chrome's New Tab Page and Side Panel with strict security containment.
Research & Sourced Reports
A hub exploring topics through sourced research reports and interactive experiment workbenches. Covers decision models, probabilistic exploration, and empirical software investigations.
Browser AI
Interactive demos of AI models that run locally in your browser, with real controls, diagnostics, and explainers.
Research
An evidence-driven audit of how the world’s most-used sites apply modern web principles.
Browser AI
Explore and compare image embeddings directly in the browser.
Web research
An investigation into links, citations, and how information moves across the modern web.
AI experiment
Does an opaque URL alone steer an LLM’s output—and does that effect track its knowledge cutoff?
Older tools and experiments that are still available.
Experiments on the future of the web: media types, navigation, and Chrome as an agent platform.
Experiments, prototypes, and tools from the archive whose deployments still work.