Skip to content
Kitbay
Skill

performance

Apache-2.0Updated on Nov 11, 202686,197GitHub stars

A contributor skill in the RTK repository for keeping the RTK command-line tool fast and small.

What the skill teaches

A contributor skill in the RTK repository for keeping the RTK command-line tool fast and small. It sets targets for startup time under 10 milliseconds, memory under 5 MB, token savings of at least 20 percent per output filter and a stripped binary under 5 MB, and has the agent record a baseline with hyperfine, the time utility and the binary size, make the change, rebuild and compare. Regressions are investigated with flame graphs, heap profiling or the token-savings tests. It lists common causes, such as regular expressions recompiled on every call, needless string copies, configuration read at startup and unused crate features, and suggests a pre-commit check and a CI job.

Capabilities

Not stated in its documentation