Cinderella-Man

Cinderella-Man / credence

Public

Credence is a semantic linter for Elixir

12
1
100% credibility
Found May 01, 2026 at 12 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Elixir
AI Summary

Credence reviews AI-generated Elixir code to find and suggest fixes for inefficient or non-idiomatic patterns.

How It Works

1
šŸ” Discover Credence

You hear about a helpful tool that spots common slip-ups in code made by AI assistants, making it more natural for Elixir projects.

2
šŸ“¦ Add to your project

You easily include it in your development setup so it's ready whenever you need to check code.

3
āœØ Check your code

You paste in the AI-generated code and run a quick review to see what could be smoother or faster.

4
šŸ’” Get friendly tips

It shows clear suggestions on better ways to write things, like shortcuts for common patterns.

5
šŸ”§ Make improvements

You apply the tips to tweak your code, feeling confident it's now more efficient and readable.

šŸŽ‰ Perfect code ready

Your code is now clean, idiomatic, and performs great, saving time and headaches down the road.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 12 to 12 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is credence?

Credence is a semantic linter for Elixir that scans code for non-idiomatic patterns and performance traps, like inefficient list traversals or Python-style habits that LLMs spit out. It takes a code string, parses it to AST, and returns issues with fixes ā€“ think Credo for style, but for meaning. Elixir devs drop it as a dev dep to validate LLM-generated modules, ensuring they run idiomatically without full compiles or tests.

Why is it gaining traction?

Unlike Credo or Dialyzer, Credence zeroes in on LLM-specific blunders ā€“ Enum.count for length, sort-then-reverse, grapheme chunking bloat ā€“ with messages that double as retry prompts. Its barebones API fits code pipelines seamlessly, catching O(nĀ²) bugs syntax tools miss. As Elixir LLM workflows explode, it's the quick semantic check devs crave.

Who should use this?

Elixir backend devs generating code via LLMs for Phoenix apps or pipelines. Ideal for teams auto-building reducers, string processors, or recursive helpers where AI fumbles idioms ā€“ lint before merge to squash inefficiencies early.

Verdict

Promising niche tool at v0.1 with 12 stars and 1.0% credibility score; docs shine, rules target real pain, but it's early ā€“ expand coverage before prod. Try it for LLM Elixir validation; mature codebases can wait.

(178 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.