#Gemini

An AI system that automatically reviews Linux kernel patches. Detected 53.6% of known bugs that passed human review. Rust implementation supports both Gemini and Claude.

Design and implementation of Kana Chat, a personal AI agent system that wraps official CLIs. Covers the tmux bridge, context isolation, and tool approval gate that make it safe to run in your own environment.

Gemini auto-generates images when you only ask about one or request a text prompt. The Saved info rule that stops it, a conversation-level fix, and where Google's fixes currently stand.

The Web Speech API + Gemini + VOICEVOX setup is complete — an AI character you can actually have a voice conversation with. Key implementation notes and impressions.

Comparing how to give AI characters personality across the Gemini, Claude, and OpenAI APIs — covering System Prompts, Prefill, and structured output.

Aiming for a characterful AI with an avatar and voice chat, I started by comparing voice APIs.

A reference sheet of camera angles, shot sizes, and camera effects for AI image generation — prompts alongside output images.

Planning a Japanese voice chat system built on the voice-chat project, using Gemini 2.0 Flash for LLM and VOICEVOX for TTS.

To stabilize body shape in Gemini Gem and Flow, I built a body reference image that includes figure information.

Testing the image generation features of Flow, now available in Google AI Pro. Findings on the He/She pronoun issue, the effectiveness of natural English prompts, and how to use Flow vs Gem.

A Gemini Gem that generated characters perfectly yesterday suddenly fell apart. A record of reviving it through prompt revision.

After reporting that Nano Banana Pro wasn't available even with a Gemini Pro subscription, it finally rolled out — here's what character-consistent image generation actually looks like in practice.