01Links
Claude Code
Put your agent inside the repo, where it can read the plan, run the skills and check its work.
Give the writing to the model I prefer for copy, while Claude checks every claim against its source.
How I use it
I think Codex writes better copy, so I give it a full brief and have Claude check every claim against the sources. I want the words to hold someone’s attention, but that doesn’t make a fluent sentence true. The writing and the checking have different jobs.
I use that split for new site copy, including these words. The defer-to-codex skill carries the brief into the CLI and keeps the output in scratch until review, so Claude applies the checked result. Codex can also use the shared skills through .agents/skills/.
How I use it: Let one model write and another check →01Links
Put your agent inside the repo, where it can read the plan, run the skills and check its work.
02Links
Keep separate jobs open across Claude accounts, with Notch showing which account has room to work.
03Links
Speak the job, the look and the stack in one brief, so your agent gets the context together.
05Links
Give apps and tools their own shortcuts, so you can switch contexts from the keyboard.
06Links
Turn Caps Lock into all four modifiers, so new shortcuts don’t clash with app defaults.
I’ll send you one build log a week: what I added to the library and what I shipped with it. You can unsubscribe any time.