# Knowledge base

Measured results from building SteerSpec, for engineers. Each article states its sample and its limits next to the numbers, and links the code that produced them at a tag or commit, so you can rerun it.

strspc is open source under the Apache-2.0 licence (https://github.com/SteerSpec/strspc-CLI/blob/v1.40.0/LICENSE). These articles report one-off measurements on this project's own rules and code. They show a pattern in that setting, not a guarantee for yours, and an article is not updated when a later release changes the behaviour it measured.

## Which model and effort level should the CI rule check use?
We measured Sonnet 5.5 and three Haiku 5.5 effort levels on the CI rule check. The smaller model at its default effort got 54 of 54 known-answer runs right against 51 of 54, at about 1/18 of the cost per CI run, and raising the effort made it flag more violations. Small sample, one repository. For: Engineers who run an LLM as a judge in CI. Published: 2026-11-10. Evidence: https://github.com/SteerSpec/strspc-manager/blob/v1.50.0/src/ruleeval/claude/bench_test.go. https://steerspec.dev/kb/ci-rule-check-model.md

Where next: https://steerspec.dev/how-to.md, https://steerspec.dev/concepts.md, https://steerspec.dev/spec.md
