I'm never sure whether to believe llms about themselves.
Anthropic create developer containers with a selection of tools installed. Is that a better guide? Some of the config seems aimed at human developers but if claude likes the tools why aren't they in this list?
Interesting - I've been digging into Claude Code's behavior patterns too, but from the cost side. The variance in how many API calls a task takes is wild. A simple refactoring can be 3 calls or 30 depending on codebase state. Curious what tools it suggested — did any of them relate to reducing unnecessary loops?
I'm never sure whether to believe llms about themselves.
Anthropic create developer containers with a selection of tools installed. Is that a better guide? Some of the config seems aimed at human developers but if claude likes the tools why aren't they in this list?
https://github.com/anthropics/claude-code/blob/main/.devcont...
Interesting - I've been digging into Claude Code's behavior patterns too, but from the cost side. The variance in how many API calls a task takes is wild. A simple refactoring can be 3 calls or 30 depending on codebase state. Curious what tools it suggested — did any of them relate to reducing unnecessary loops?
I made tilth much for the same reasons. Its MIT licence, open source and saves you 30% token use (benchmarked)
https://github.com/jahala/tilth
Interesting about the `semgrep` tool; I've never heard of it before.
The strange thing (or not) is that I have got completely different set of recommendations coming from claude.
Also; funny thing is I use mise a lot more than brew and claude could pick up on that.