Agent Skill
2/7/2026

explain

Explain all bluera-base plugin functionality in human-readable format

B
blueraai
1GitHub Stars
1Views
npx skills add blueraai/bluera-base

SKILL.md

Nameexplain
DescriptionExplain all bluera-base plugin functionality in human-readable format

name: explain description: Explain all bluera-base plugin functionality in human-readable format argument-hint: "[overview|features|commands|behaviors|config|philosophy]" allowed-tools: [Read, Write, Edit, Bash, Glob, Grep, AskUserQuestion]

bluera-base Explained

User guide for the bluera-base plugin.

Subcommands

  • /bluera-base:explain or /bluera-base:explain all - Show everything
  • /bluera-base:explain overview - What is bluera-base?
  • /bluera-base:explain features - Configurable features
  • /bluera-base:explain commands - Available commands
  • /bluera-base:explain behaviors - Automatic behaviors
  • /bluera-base:explain config - Configuration system
  • /bluera-base:explain philosophy - Design principles

Algorithm

Present the documentation below to the user. This is a documentation command - output the content directly, don't just acknowledge it.


Overview

What is Bluera Base?

Bluera Base is a conventions plugin for Claude Code. It provides shared development workflows, quality gates, and patterns that you can use across all your projects. Instead of copying hooks and skills between repositories, install Bluera Base once and every project gets the same standards.

What It Does

Enforces quality patterns:

  • Blocks anti-patterns like any types, fallback code, and commented-out code
  • Requires atomic commits with conventional format
  • Prevents manual versioning (use the release workflow instead)

Provides development workflows:

  • Atomic commit creation with README/CLAUDE.md awareness
  • Multi-agent code review
  • Iterative development loops (milhouse)
  • Automated release cutting with CI monitoring

Automates repetitive tasks:

  • Desktop notifications when Claude needs input
  • Auto-commit on session end
  • Duplicate code detection
  • Lint and typecheck validation after edits

Supported Languages

Bluera Base works with 13 languages:

JavaScript/TypeScript, Python, Rust, Go, Java, Kotlin, Ruby, PHP, C#/.NET, Swift, Elixir, C/C++, Scala


Quick Start

Install

# Add the Bluera marketplace (one-time)
/plugin marketplace add blueraai/bluera-marketplace

# Install the plugin
/plugin install bluera-base@bluera

Configure

# Interactive setup - walks through each feature
/bluera-base:settings init

# Or enable features individually
/bluera-base:settings enable notifications
/bluera-base:settings enable strict-typing

Explore

# See all commands
/bluera-base:help

# See this documentation
/bluera-base:explain

Features

Features are opt-in capabilities you can enable or disable. See references/features-guide.md for detailed descriptions, usage examples, and platform support.

  • Desktop Notifications -- Alerts when Claude needs input (default: ON)
  • Auto-Learn -- Track command patterns, suggest CLAUDE.md updates
  • Auto-Commit -- Prompt to commit on session end
  • DRY Check -- Detect duplicate/copy-paste code
  • Strict Typing -- Block unsafe type patterns (any, as, cast)
  • Milhouse Loop -- Iterative development with gate commands
  • Deep Learn -- Semantic session analysis via Claude CLI
  • Coverage -- Enforce minimum test coverage thresholds

Commands

See references/commands-reference.md for the complete commands table organized by category.

Automatic Behaviors

See references/automatic-behaviors.md for details on anti-pattern detection, release protection, auto-validation, context preservation, and configuration.


Philosophy

Bluera Base is opinionated. Here's why.

Fail Fast

Principle: Errors should be visible immediately, not hidden.

When something goes wrong, you should know about it right away. Silent failures lead to bugs discovered much later when they're harder to fix. Bluera Base blocks problematic patterns at edit time rather than letting them slip through.

No Fallbacks

Principle: Code works as designed or fails visibly.

Fallback code, graceful degradation, and "just in case" defaults often hide bugs. If your code expects a value to exist, it should fail if it doesn't - not silently use a default that may or may not be correct. This makes bugs obvious instead of subtle.

Atomic Commits

Principle: Each commit is one logical change.

A commit should do one thing. "Fix login bug" is good. "Fix login bug and update README and refactor utils" is three commits. Atomic commits make git history useful for understanding what changed and why, and make reverts safe.

Clean Code

Principle: If it's in the codebase, it runs.

No commented-out code "for reference." No deprecated functions kept "just in case." No unused exports. Dead code is noise that makes the codebase harder to understand. If you need old code, that's what git history is for.

Strict Typing

Principle: Types are documentation that the compiler checks.

any and as casts defeat the purpose of TypeScript. They tell the compiler "trust me" when you should be telling it what to expect. Strict typing catches bugs at write time instead of runtime, and makes refactoring safe.

Skills Info
Original Name:explainAuthor:blueraai