Your agent just got busted.

SlopCop checks the code your AI writes.

It finds the bugs, security holes and waste in every change. Your agent explains what’s wrong in plain English, and fixes it.

Runs in
Claude Code
Codex
GitHub PRs

Also Cursor, Lovable and Bolt, and other agents.

A detective in a trench coat and fedora, seen from behind, shines a torch across a dark hall where ranks of identical tin robots sit typing. The robot in the beam has looked up, and its eyes are red.

What SlopCop catches.

An API key your AI left in the code
A secret is written into the source, where anyone with the code can read it.
A search box that leaks your data
What a visitor types ends up inside a database query.
A failed save that looks like a success
An error is caught and thrown away, so nothing reports the failure.
A page that runs up your database bill
It asks the database once for every row it shows, instead of once for the page.

See more

The problem isn’t always in the file you changed.

Something a visitor types on one page can do harm three files away. SlopCop follows your data across your whole app and shows your agent where it goes wrong.

routes.pywhere it comes in
return find_user(request.args["name"])
users.pywhere it is passed on
return run_query(FIND_USER % name)
db.pywhere it does harm
return conn.execute(sql).fetchall()

SlopCop reports one finding

A request parameter reached the query argument of Connection.execute(sql)
A detective in a hat and trench coat, seen from behind, stands back with one hand in his pocket and holds a magnifying glass up in front of him.

Fixed while your agent is still on the case.

Connect SlopCop to your coding agent. When a check finds a problem, it goes straight back to your agent, which tells you what’s wrong in plain English, fixes it and checks again. You’re never the middleman.

  1. Writeusers.py
    return run_query(FIND_USER % name)
  2. Check$ slopcop check
    db.py:7 a request parameter reached the query
  3. Fixusers.py, db.py
    conn.execute(sql, (name,))
  4. Recheck$ slopcop check
    No new findings | Analysis complete

On every pull request too

A riveted tin robot types at a desk under a lamp. A detective in a hat and trench coat leans over its shoulder, one hand on it, reading the page in the typewriter.

Thirteen languages.

  • C
  • C++
  • C#
  • Go
  • Java
  • JavaScript
  • Kotlin
  • PHP
  • Python
  • Ruby
  • Rust
  • Scala
  • TypeScript

Coverage differs by language. Repositories on GitHub, GitLab.com and Bitbucket Cloud.

It isn’t another AI checking the AI.

SlopCop checks your code against a rulebook, so it doesn’t guess. The rulebook runs on Bifrost, our open-source static analysis engine, and there is no AI model in it.

Same code, same answer
Run the check twice and you get the same findings.
Never a false all-clear
A check that could not finish says incomplete, never clean.
Only what the agent changed
Problems already in your code are counted, and never fail the check.
Every finding shows its proof
The rule, the file, the line, and how sure SlopCop is of the match.
A police line-up of five identical riveted tin robots against a height chart. One has been made to step forward.

Safe to point at private code.

Read-only
It reads the repositories you choose. It cannot change them.
Never runs your code
Nothing is built and nothing is executed.
Checked by rules, not an AI
One check is the exception: to spot duplicated code, it sends your functions to an AI model.

What leaves your machine

A police sergeant stands guard beside the barred, padlocked door of an evidence room. Shelves of file boxes show through the bars.

Case notes.

Read the blog
  1. Jonathan Ellis

    A cheapskate's guide to coding with expensive models

    It seems straightforward that you should be able to save money by having a smart, expensive model delegate straightforward tasks to cheaper models, and this is in fact the case. With Mjolnir , you can even do this cross-provider; here's price vs performance o…

    1. Jonathan Ellis

      Comparing Jev, Gemini, and OpenAI decision models

      One of the ways that Mjolnir makes your agents smarter is by delivering relevant github events to them so they stay looped in to what's happening with the rest of the team. Mj github notifications Some of these can be done deterministically: activity on your…

    2. David Baker Effendi

      Code intelligence has escaped the code

      Bifrost v0.14.4 correlates Java configuration reads with entries in named .properties files, and reports ambiguous or incomplete matches as such.

Rules for your own codebase.

Every codebase has mistakes that keep coming back. Turn yours into rules, and your agent stops making them.

Written by you or your agent
A rule is a file in your repository, so it is reviewed and versioned like the rest of your code. Build your own rules
Run with every check
The next check picks a new rule up and runs it beside ours. There is nothing else to set up.
Shared across your team
On GitHub, rules kept in one shared repository also run in your organization’s other repositories. Private rules run only in private ones.

We can write them for you

Tell us the mistakes your team keeps fixing by hand, and we write the rules and set them up with you. How custom rules work

Create your account.

Your first 1,000 checks each month are free. You’ll set up your coding agent next.

or

At least 15 characters. We’ll email you a code to confirm the address.

By continuing, you agree to the Terms of Service and acknowledge the Privacy Policy.

Have an account?

Set it up with us.

Leave your email and we’ll arrange a call about running SlopCop in your team’s coding agents and CI, or anything else you’d like to know.

We’ll use your email to respond.