FrameCue: building a video workflow that keeps getting more useful
The decisions behind FrameCue: isolated video projects, reusable assets and effects, shared local tools, targeted revisions, and recoverable production records.
Read articleWRITING
Posts
Project stories and practical notes, from the problem to the implementation and the choices along the way.
game-creator: images, video, and audio in an agent workflow
How game-creator 2.0 organizes media requests, resumable jobs, and output validation for AI agents.
Read articlePulseGate: a desktop view of AI activity
A Windows floating island and dashboard for task reminders, token usage, known costs, and system resources.
Read articleMaintaining an AI-assisted knowledge base with clear boundaries
How siyan-brain-skills organizes an Obsidian second brain with source tracking, layered knowledge, and bounded AI edits.
Read articleConnecting visual input to an agent with vision-bridge
A Python-based bridge for image understanding, OCR, and video analysis when an agent lacks a visual input channel.
Read articleBringing 100,000 Particles to the Browser: Vibe Particles Web
A creative-coding retrospective on bringing GPGPU particles, audio, MIDI, and recording into the browser.
Read articleWhy AI Needs a Reviewer Willing to Tell You the Uncomfortable Truth
What diss-me taught me about turning sharp criticism into an AI workflow that helps rough ideas become buildable plans.
Read article