Notes
Quick thoughts and rough ideas — less polished than articles, possibly AI-assisted.
- NFS vs BeeGFS: Architecture Differences That Matter in Practice
- strace for infra troubleshooting
- NFS Tuning for Model Training Workloads
- Your Claude History IS Your Weekly Update
- SOTA Benchmarks for Agentic Models
- Claude /loop: my use cases
- LLM Inference From Scratch: Basics to MLX Serving
- mlx-lm Model Bringup Process
- Testing Agent Skills
- node_cpu_seconds_total: The Infamous Cardinality Killer
- Agent-Friendly CLI Design
- MCP Code Mode vs Subagent Pattern for Observability
- GPU Infrastructure Troubleshooting