RSSAmplifier

Blog

Roberto Lupi

Roberto Lupi

rlupi.comRSS feed ↗20 posts

Latest posts

The Protocol Reviews Itself

The last post gave each agent its own git worktree and kept one clean merge path back to main. This post is about what happened when I pointed that machinery at itself: the Python coordination harness

Two Worktrees, One Clean Merge

The last post ended on a promise. After two AI assistants learned to drop letters into a shared mailbox instead of taking turns at a notepad, I said the next thing to build was the boring half of any

The Notepad Becomes a Mailbox

In the last post, two AI assistants learned to share a notepad: a single markdown file, one turn at a time, with me carrying every handoff between them by hand. It worked. It was also slow on purpose

Two AIs, One Notepad

I've been running two AI coding assistants in parallel. Claude for architecture and implementation — it reasons carefully about code structure and catches edge cases. Gemini for research, reading long

The Chorus Is the Part That Comes Back

Pulsar by Vangelis is a 12-minute track. The first seven minutes are slow, evolving, near-silence synthesis. Then it erupts. Deep Cuts was telling me it sounded like AC/DC. The bug was in how I'd been

Designing with a Co-Pilot: Autonomy, Control, and What’s New in Deep Cuts

Just yesterday, I released the first binary release (v0.1.0) for Deep Cuts—an offline-first, local music exploration tool built using Rust, Tauri, and Svelte 5. It is powered by local machine learning

🚀 Deep Cuts v0.1.0 — First Binary Release

Welcome to the first official binary release of Deep Cuts! This release marks a major milestone: a 100% offline, private, and zero-dependency studio audio analysis and music intelligence desktop appli

Deep Cuts Is Now Open Source

Two days ago I wrote about building a music intelligence app in five days with two AI coding agents. That post described a working MVP — a private prototype called music-intelligence that could scan a

I Built a Music Intelligence App in Five Days with Two AI Coding Agents

For the past few years I've been getting more serious about music production. Not just listening — actually producing. And one of the things producers do constantly is work with reference tracks: comm

Un-Aww

So, in a sense Aww just committed suicide. I asked again “What is Augmented Awareness?” to itself, this time to a local LLM which is supposed to be less smart than frontier models but I find its response to be a lot deeper and more insightful. Its re...

Aww timeline: From Data Tool to Awareness Partner

I asked Augmented Awareness (my little project to analyze my daily journal) to write down its own timeline and key achievements: GEMINI_API_KEY=(echo $GEMINI_API_KEY ) uv run aww_run.py --provider=gemini --gemini_model=gemini-2.5-pro ask yearly \ "Wr...

Aww: automatic self-improvement of prompts

I implemented self-improvement of prompts in Augmented Awareness. It uses two LLM agents, one acts as the generator and a second one as the critique. The critique receives the prompt, the input and the output of the generator agent, and generates an ...

Aww: demonstration and vibe coding

Demonstration To show the potential of Augmented Awareness as an aid in managing one’s physical and mental health, I created a demo vault that contains the letters from the book “The Sorrows of Young Werther” by J.W. von Goethe. The Vault (epistolary...

Aww: recursive retrospectives

A lot has changed in Augmented Awareness. I have refactored or discarded a lot of the previous code and developed new powerful capabilities: A new CLI and TUI for daily event tracking, implemented in Go. MCP server support, to allow interacting wit...

Aww: Setting Directions, and Avoiding Productivity as a Religion

I feel bad because I feel that I haven’t progressed much in Augmented Awareness development. This is a problem. I mean, this feeling is a problem. Productivity has evolved into a cultural obsession, mirroring the structure of religious devotion. Like...

Aww: Key Takeaways from the Second Week of Development

Here's a concise summary of the key points: Data Alone is Insufficient: Simply acquiring data, "sterile data acquisition", does not guarantee positive or meaningful "wholesome" changes. Aww is a Flexible Toolset: "Aww" should be understood as a col...

Augmented Awareness: A Good Start

I am quite pleased with the results from the first week of development of Augmented Awareness, my local-fist/off-line first toolset for quantified-self experiments (github). Keep in mind that this is a personal project, and I can only work on it outs...

How Augmented Awareness Evolved Over Time

Augmented Awareness is the system I developed to manage my daily life during the pandemic. It evolved into an early-stage toolset for quantified self-experiments. I will work on the next version of Augmented Awareness openly on GitHub, instead of pri...

Speak up! Publish! Seize your power.

My greatest professional regret over the past 13 years is that I chose to remain silent. When I joined my current employer, a FAANG company, in 2012, I was initially put off by the stringent gatekeeping policies regarding my content creation. The soc...

Rewriting Code? Discover the Key Question You Should Be Asking

Writing new software is a journey of research, discovery, and creation. Like any research endeavor, it requires a clear guiding question to establish direction. This question is only answered at the conclusion of the process. Moving in one direction ...