Essential AI-augmented utilities for development - commands, agents, and hooks following Unix coreutils philosophy
This plugin is not yet in any themed marketplace. To install it, you'll need to add it from GitHub directly.
Choose your preferred installation method below
A marketplace is a collection of plugins. Every plugin gets an auto-generated marketplace JSON for individual installation, plus inclusion in category and themed collections. Add a marketplace once (step 1), then install any plugin from it (step 2).
One-time setup for access to all plugins
When to use: If you plan to install multiple plugins now or later
Step 1: Add the marketplace (one-time)
/plugin marketplace add https://claudepluginhub.com/marketplaces/all.jsonRun this once to access all plugins
Step 2: Install this plugin
/plugin install ai-coreutils@allUse this plugin's auto-generated marketplace JSON for individual installation
When to use: If you only want to try this specific plugin
Step 1: Add this plugin's marketplace
/plugin marketplace add https://claudepluginhub.com/marketplaces/plugins/ai-coreutils.jsonStep 2: Install the plugin
/plugin install ai-coreutils@ai-coreutilsWhat GNU coreutils are to Unix, ai-coreutils are to AI-assisted development
Essential AI-augmented utilities for development following Unix coreutils philosophy.
Unix coreutils (ls, grep, sed) operate on text streams through standard interfaces. ai-coreutils leverage Claude's capabilities through slash commands, agents, hooks, and MCP servers.
Traditional: User → Unix Tools (ls, grep, sed)
AI-Enhanced: User → Claude → Unix Tools + AI Capabilities
Claude sits between user and tools, leveraging both traditional Unix utilities and AI-specific operations. Not replacing, augmenting.
| Unix System | Description | Claude Code System | Description |
|---|---|---|---|
Coreutils (ls, grep) | Small C programsOperate on text | Slash Commands (/extract, /summarize) | Prompts guiding Claude's tool usePredictable text operations |
Shell/Pipes (bash, |) | Composition layerChain operations | Hooks (PostToolUse, SessionStart) | Event-driven automationOrchestrate Claude's actions |
Daemons (sshd, httpd) | Single-purpose servicesAlways available | Agents (doc-writer, test-generator) | Focused Claude instancesSpecialized tasks |
System Services (systemd, cron) | Complex coordinationInter-process comm | MCP Servers (Model Context Protocol) | External tool integrationState & communication |
Unix: stdin/stdout (text streams) Claude Code: tool calls/results + conversation context
Claude sits between you and Unix tools, adding semantic understanding:
| Traditional | AI-Augmented |
|---|---|
grep -rn "TODO" . → You read and prioritize manually | /extract TODO → Claude provides context, priority, explanation |
head -n 20 file.py → You interpret what code does | /summarize file.py → Claude explains purpose, components, dependencies |
diff old.py new.py → You analyze if equivalent | /compare old.py new.py → Claude identifies semantic equivalence and risks |
# Add marketplace
/plugin marketplace add Piotr1215/aiverse
# Install plugin
/plugin install ai-coreutils@aiverse
# Verify installation
/plugin list
Prompts that guide Claude to perform specific operations with consistent output format. Claude uses its tool capabilities under the hood. Examples: extract structured data, summarize in bullet format, semantic comparison.
Pre-configured Claude instances with focused behavior and limited scope. Each agent has specific system prompts and tool access. Examples: documentation writer, test generator, security reviewer.
Bash/Python scripts triggered by Claude Code lifecycle events (SessionStart, PostToolUse, Stop). Orchestrate workflows, call external systems, automate responses to Claude's actions.
Model Context Protocol servers that extend Claude's capabilities. Provide custom tools, maintain state, enable communication between instances, integrate external systems.
Extract patterns from code with semantic context and priority analysis.
Example:
/extract TODO src/
Output: Structured list with context, priority (High/Medium/Low), and explanation of why each match matters.
Condense file to essential information with language-specific understanding.
Example:
/summarize src/auth.py
Output: Purpose statement, key components, dependencies, notable issues.
Compare files for semantic equivalence, not just syntax differences.
Example:
/compare old_version.py new_version.py
Output: Equivalence assessment, functional differences, risk level, recommendation.
Technical documentation specialist for API docs and READMEs.
Tools: Read, Write, Grep, WebFetch Focus: Clear, comprehensive documentation with runnable examples
Example:
@doc-writer document the authentication module in src/auth.py
Output: Standard API documentation with signatures, parameters, returns, examples, error cases.
Test case generation expert focusing on coverage and edge cases.
Tools: Read, Write, Grep Focus: Comprehensive test suites achieving >80% coverage
Example:
@test-generator create tests for src/auth.py
Output: Runnable test file with happy path, edge cases, error cases, integration tests.
Automatically sends notification when Claude finishes work (if you're not watching).
Event: Stop Behavior: Detects if you're in active tmux window, sends platform-appropriate notification Platforms: Linux (notify-send), macOS (osascript)
No configuration needed - automatically enabled on installation.
/extract TODO src/
Output includes high-priority security TODOs first, explains why each matters.
/compare auth_v1.py auth_v2.py
Claude identifies that files are "mostly equivalent" despite async refactoring, highlights added rate limiting as enhancement, warns about missing password validation.
@doc-writer document src/auth.py
Creates comprehensive API documentation with runnable examples.
ai-coreutils includes automated tests to ensure quality and prevent regressions.
# Validate plugin structure
bash scripts/validate_plugin.sh
# Run hook tests
bats tests/notify_test.bats
npm install -g batsapt-get install jq or brew install jqgit checkout -b feature/amazing-feature)scripts/validate_plugin.sh passesgit commit -m 'Add amazing feature')git push origin feature/amazing-feature)MIT - See LICENSE file for details.
0.1.0