AI & LLM
Claude Code /goal: Make Completion a Deterministic Gate
Jason Yang · 22 Jul, 2026
Claude Code's /goal decides it's done from the worker's self-report. Add a Stop hook that verifies completion with exit codes, not the model's word.
Blog posts — page 17
AI & LLM
Running Auto Research in Claude Code — A Bash-Driven Experiment Loop
Jason Yang · 22 Jul, 2026
Karpathy's autoresearch pattern rebuilt for Claude Code: a Bash loop that scores each AI-tried hypothesis, keeps improvements, and rolls back failures with git.
Tools
Obsidian Local Image Attachments: Why I Ditched Imgur
Jason Yang · 19 Jul, 2026
Imgur's OAuth flow stopped working, so I moved my Obsidian attachments local: the /assets + UUID setup, the migration, and the settings that cost me an evening.
AI & LLM
Three Claude Code Skills for a Subtitle Workflow
Jason Yang · 12 Jul, 2026
Three Claude Code skills for downloading YouTube captions as SRT, transcribing audio with a local Whisper server, and translating embedded subtitle tracks.
AI & LLM
Harness Engineering: Keeping AI From Running Wild
Jason Yang · 09 Jul, 2026
What an AI harness is and how to build one — validation, gates, and Generator ≠ Evaluator separation at each stage of the SDLC.