Get This Tool
Stupify
Pricing
- Model
- Usage-Based
Summary
AI-generated code passes the linter, passes the skim, and then quietly rots — because slop isn't a syntax error, it's a judgment call your static analysis tools are never going to make.
stupify runs PR reviews against a corpus of code you actually respect — your own best files, or commit-pinned exemplar packs from coders like dtolnay, DHH, or Rich Harris — and names what's wrong in terms of concrete primitives, not style opinions. It runs on your own Codex plan, so there are no additional API keys or servers to manage. The rubric keeps findings small and actionable: a named helper that got dissolved into its call sites, a hand-rolled state machine where a library call already exists, a duplicated data source already drifting from its canonical version. Once findings are addressed, it posts one line and stops. What it cannot do is catch slop that isn't representable in the corpus — if your taste reference doesn't cover a pattern, neither does the review.
Bottom line: Pick stupify when you have a code corpus worth enforcing and a PR workflow worth protecting; reconsider when your team has no existing reference code and no appetite to curate one.
Community Performance Report Card
No community ratings yet. Be the first to rate this tool!
Community Benchmarks Community
Sign in to submit a benchmarkNo community benchmarks yet. Be the first to share a real-world data point.
Pros
Sign in to edit- Corpus-grounded reviews rather than abstract rule sets, which means findings reference actual code decisions rather than style preferences no one can trace back to a reason.
- Curated taste packs from named, commit-pinned coders, so teams without a large internal corpus can borrow a coherent standard from someone whose output they would point a new hire at.
- Runs on your own Codex plan with no additional API keys or servers, which means there is no third-party service holding your diffs and no separate billing surface to manage.
- MIT licensed and self-hostable, so the review pipeline stays inside your infrastructure and there is no vendor dependency to negotiate around.
- Posts one closing line once findings are addressed and goes quiet — the opposite of a bot that re-flags resolved issues on every subsequent push, which means the PR thread stays readable.
Cons
Sign in to edit- The review quality ceiling is the quality of the corpus: patterns not represented in your reference files produce no finding, which means AI slop in domains your corpus does not cover ships undetected — teams hit this when a new language or framework appears in contributions before it appears in their exemplar files.
- No API and no programmatic output surface means stupify cannot be wired into a CI gate that blocks a merge automatically; teams that need a hard merge block on review findings route through a different tool or build a wrapper themselves.
- Taste packs from external coders are concrete and commit-pinned, but a rubric derived from a codebase in a different domain or language than yours will generate findings that do not transfer — teams that adopt a pack without auditing it first report spending more time dismissing irrelevant findings than addressing real slop, and those teams migrate to maintaining their own corpus or to a general-purpose code review tool with configurable rules.
Community Reviews
Sign in to write a reviewNo reviews yet. Be the first to share your experience.
About
- Platforms
- CLI via npm
- API Available
- No
- Self-Hosted
- Yes
- Last Updated
- 2026-06-25T06:17:46.860Z
Best For
Who it's for
- Developers maintaining high code quality standards
- Teams avoiding AI slop in contributions
- Users with existing code corpora or preferred style references
What it does well
- Reviewing PRs for AI-generated code patterns
- Enforcing standards from specific respected codebases
- Providing actionable fixes for slop in code contributions
- Running automated reviews on own model plan
Integrations
Discussion Community
Sign in to commentNo discussion yet. Sign in to start the conversation.
Spotted incorrect or missing data? Join our community of contributors.
Sign Up to ContributeCommunity Notes & Tips Community
Sign in to contributeBe the first to contribute. General notes, observations, gotchas, and tips from people who use this tool day-to-day.
Frequently Asked Questions
- Is Stupify free?
- Yes — Stupify is fully free to use. There is no paid tier.
- Is Stupify open source?
- Yes. Stupify is open source.
- Can I self-host Stupify?
- Yes. Stupify supports self-hosting on your own infrastructure.
- What platforms does Stupify support?
- Stupify is available on: CLI via npm.
Hours Saved & ROI Stories Community
Sign in to contributeBe the first to contribute. Concrete time/cost savings, with context. e.g. "Cut my code review backlog from 4h to 45m per week."
Curated lists that include this category
Most linters operate on rules that are satisfied the moment the code compiles correctly. The judgment that a helper should not have been inlined, that three boolean flags are doing the work of a library call, or that a duplicated phrase list is already out of sync with its canonical source — none of that surfaces in a lint pass. stupify exists in that gap. The core workflow is one command: `npx @stupify/cli` primes Claude Code or opens a PR for review, and the tool works through the diff against a corpus of code you designate as the standard. Each finding names the slop and names the primitive to use instead.
The differentiating mechanism is the corpus. Instead of encoding taste as abstract rules, stupify grounds every finding in actual code — your own best files, or curated ‘taste packs’ of commit-pinned exemplar files from coders like antirez, Sindre Sorhus, Mitchell Hashimoto, and others listed on the project site. A rubric sits on top of the corpus to define what counts as slop and to keep fixes scoped. The vendor states that because reviews run on your own Codex plan, a careful review costs a fraction of a raw API call.
stupify fits teams with an existing code culture they want to enforce — maintainers who have a clear sense of what their codebase should look like and contributors submitting AI-assisted PRs that technically pass review but carry the texture of generated filler. It does not fit teams that have no reference corpus and no one willing to build one: the taste packs help, but a borrowed rubric from a coder you have never read is a shaky foundation for blocking contributions. The tool is MIT licensed, self-hostable, and requires no separate server infrastructure — all compute runs through the user’s own Codex plan.