What a check finds.

On the lines it wrote

A check analyzes the files your agent’s change writes and reports the findings on the lines it added or touched. Findings already in the code are counted, and never fail the check.

Traced to a rule

Each finding names the rule that raised it, its severity, its certainty, and the file and line, from Bifrost’s static analysis rather than a language model. The docs list the kinds of findings a check reports.

Never a false all-clear

A rule that can’t finish makes the check incomplete, never clean, so your agent knows what wasn’t covered. The guide to reading a report says what each status means.

In your agent’s loop

Your coding agent runs slopcop check after each edit and fixes what it finds before you push. It ignores a false positive with slopcop ignore, and later checks leave that finding out. The docs show how to set SlopCop up in your coding agent.

On every pull request

The same check runs on each push to a GitHub pull request, with every new finding on its line in Files changed. See pull request checks.

Questions.

Does a clean check mean the code has no bugs?

No. A clean check means every rule the check ran completed for the files the change writes, and no new finding failed the check. It is a claim about those rules, files and lines, and not about the rest of the code.

Does SlopCop fix the code?

No. SlopCop reports each finding to your coding agent. The agent makes the fix and runs the check again, until no new findings are left.

Which languages does a check cover?

Thirteen: C, C++, C#, Go, Java, JavaScript, Kotlin, PHP, Python, Ruby, Rust, Scala and TypeScript. Coverage differs by language and by rule: each rule runs only on the languages it has been validated on.

What does a check cost?

A workspace gets 1,000 free checks each calendar month. Each check after those costs $0.001, paid from prepaid credits, and repositories are free. Pricing lists the credit packs and says what happens when credits run out.