[CodeDeficient]

User description

Closes #103.

KarpeSlop started as a slop index: one score that says how much of a codebase is slop, and what kind. 2.0.0 dropped the 1.x point total because it grew with repository size, and the three axes went with it. This brings them back, normalized.

What changes

• Every built-in rule has an axis: noise (information utility), lies (information quality) or soul (style and taste). --help lists it next to the severity.

• Every run that scans application code ends with a score per axis and a total:

  Slop index (application code, higher is worse):
    Noise  information utility     0.6  restating_comment, production_console_log, filler_comment
    Lies   information quality     7.8  unsafe_object_assertion, unsafe_double_type_assertion, unsafe_array_assertion
    Soul   style and taste         3.5  type_assertion, non_null_assertion, any_type_usage
    Total                         11.9
  

• Each rule adds weight(severity) × ln(1 + findings per 1,000 lines) to its axis, with weights 8/4/2/1. Only findings in application source count: tests and package manifests are left out.

  • Per 1,000 lines, so repositories of different sizes compare.
  • Logarithmic, so one rule repeated hundreds of times cannot outweigh several different kinds of slop. v1's Soul score was 95% magic_css_value.

• The JSON report adds slopIndex and an axis on every issue and byType entry.

• customPatterns take an optional axis (default soul); an unknown value makes the config invalid.

• The README, package description and codemeta lead with the three axes again, and the README credits Karpathy's "slop index" idea.

Numbers

• TruckFull (210k-line Next.js app, --quiet): 11.9, as above, leaving out the Shorthand ambient module declarations cause hallucinated_import false positives #100 false positives. With them, Lies reads 20.0.
• KarpeSlop's own source: no findings, 0.

Testing

• npm test: 157 pass, 4 of them new. They cover size-independence, a repeated rule versus varied slop, axes in the report and summary, and invalid axis values.
• npm run typecheck is clean.
• Reviewed with /review (reviewer on Haiku 5.5): both commits PASS. The P2 notes from the first review are fixed in the second commit. The remaining P3 is filed as a follow-up issue.

Follow-ups: #107 (taste rules for the Soul axis), #106 (type_assertion volume).

🤖 Generated with Claude Code

---

CodeAnt-AI Description

Score code findings across noise, lies, and soul

What Changed

• Each scan now ends with a score for Noise, Lies, and Soul, plus a total.
• Scores use application-code findings per 1,000 lines, making results comparable across repository sizes. Test files and package-age warnings are not scored.
• The JSON report includes the scores and an axis for each finding. Custom rules can choose an axis; the default is Soul.

Impact

✅ Comparable scores across repository sizes
✅ Clearer separation of noise, misleading code, and style findings
✅ Package-age warnings stay out of code scores

💡 Usage Guide

Checking Your Pull Request

Every time you make a pull request, our system automatically looks through it. We check for security issues, mistakes in how you're setting up your infrastructure, and common code problems. We do this to make sure your changes are solid and won't cause any trouble later.

Talking to CodeAnt AI

Got a question or need a hand with something in your pull request? You can easily get in touch with CodeAnt AI right here. Just type the following in a comment on your pull request, and replace "Your question here" with whatever you want to ask:

@codeant-ai ask: Your question here

This lets you have a chat with CodeAnt AI about your pull request, making it easier to understand and improve your code.

Example

@codeant-ai ask: Can you suggest a safer alternative to storing this secret?

Preserve Org Learnings with CodeAnt

You can record team preferences so CodeAnt AI applies them in future reviews. Reply directly to the specific CodeAnt AI suggestion (in the same thread) and replace "Your feedback here" with your input:

@codeant-ai: Your feedback here

This helps CodeAnt AI learn and adapt to your team's coding style and standards.

Example

@codeant-ai: Do not flag unused imports.

Retrigger review

Ask CodeAnt AI to review the PR again, by typing:

@codeant-ai: review

Check Your Repository Health

To analyze the health of your code repository, visit our dashboard at https://app.codeant.ai. This tool helps you identify potential issues and areas for improvement in your codebase, ensuring your repository maintains high standards of code health.

[codeant-ai]

  CodeAnt Review Status

✅  Review completed

Status	Commit	Started (UTC)	Finished (UTC)
✅  Reviewed your PR	2548a62	Oct 26, 2026 · 12:00	12:04

[codeant-ai]

---

Thanks for using CodeAnt! 🎉

We're free for open-source projects. if you're enjoying it, help us grow by sharing.

Share on X ·
Reddit ·
LinkedIn

[coderabbitai]

Caution

Review failed

The pull request is closed.

Finishing Touches

📝 Generate docstrings

• Commit to this branch
• Create a new PR

🧪 Generate unit tests (beta)

• Commit to this branch
• Create a new PR

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[codeant-ai]

  CodeAnt Nitpicks

1 code suggestion

1. When a scan has application lines but no findings, generateReport returns above this line, so the promised three axis scores and total are never printed.

Logic error · ai-slop-detector.ts:1045