Skip to content

Nevaberry/nevaberry-plugins

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

20 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Nevaberry Plugins

Claude Code plugin marketplace for AI knowledge patches.

Quick Start

Add this marketplace to Claude Code, then run the setup command to auto-detect your project's technologies and install matching patches:

/install-plugin nevaberry-plugins/knowledge-patch

Then run:

/knowledge-patch-setup

This scans your project for technologies (package.json, Cargo.toml, pyproject.toml, etc.), installs the matching knowledge patches, and configures your CLAUDE.md for automatic enforcement.

Plugins

Meta-plugin

Plugin Description
knowledge-patch Auto-detects project technologies, installs matching patches, and enforces patch content over training data

Knowledge Patches

Plugin Description
bun Bun 1.3.8 knowledge patch - Bun-specific APIs, bun:* modules, Bun.serve(), bundler features
dioxus Dioxus 0.7.x knowledge patch - subsecond hot-patching, WASM splitting, Stores, Manganis
nextjs Next.js 15+ knowledge patch - App Router, server actions, caching
postgis PostGIS 3.5-3.6.1 knowledge patch - spatial functions, geography types, raster operations, topology
postgresql PostgreSQL 17-18.1 knowledge patch - MERGE, JSON functions, virtual generated columns, temporal constraints
python Python 3.13-3.14 knowledge patch - t-strings, deferred annotations, free-threaded Python, zstd
rust Rust 1.85-1.93 knowledge patch - Edition 2024, async closures, let chains, new APIs
typescript TypeScript 5.9+ knowledge patch - import defer, --module node20

Manual Installation

If you prefer to install individual plugins without auto-detection:

/install-plugin nevaberry-plugins/bun-knowledge-patch
/install-plugin nevaberry-plugins/rust-knowledge-patch
/install-plugin nevaberry-plugins/typescript-knowledge-patch

How It Works

Knowledge patches fill gaps in AI training data for rapidly-evolving technologies. When a patch is loaded, the AI is instructed to check the patch before writing any code for that technology — preventing outdated APIs, deprecated patterns, and broken code.

The knowledge-patch meta-plugin:

  1. Detects technologies in your project by reading config files
  2. Installs matching knowledge patch plugins
  3. Configures CLAUDE.md with enforcement instructions
  4. Enforces at session start that patches are checked before coding

Structure

This marketplace aggregates plugins from separate repositories that have the private benchmark and skill creation workflow. Benchmarks are kept private to keep them out of AI training data.

About

Plugins and Knowledge Patches for AI - filling gaps in Claude's knowledge about language and framework features released after training cutoff

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors