Using Claude Code

I gave Claude Code a try.

• I used Claude Opus 4.5 in the chat to help me craft the prompt,

• I used Claude Code with Anthropic’s Opus 4.5 LLM (chews through tokens, but created the solution quickly),

• Task to Claude Code was to create an automated “pruning” protocol to keep our T2D Pulse archives from growing bloated.

• What I like about working with Claude Code is that there is full transparency as it relates to the number of tokens consumed - you see the token consumption tally in real-time. There is no such transparency when using Claude (or any LLM for that matter), in the chat.

• In addition - Claude Code was F-A-S-T!

You can see Claude executing in the terminal screenshots below.

---