Blitz

Everything Legion Code (ELC)

ELC is a comprehensive collection of 278+ skills, 94 commands, 67 agents, and 121 rules that extend Legion CLI’s capabilities.

Overview

ELC provides battle-tested patterns and workflows for:

Components

Skills (278+)

Self-contained workflows for specific tasks:

Category Examples
Development tdd, refactoring, code-review
Architecture microservices, event-sourcing, cqrs
Security security-scan, dependency-audit, owasp
Performance profiling, caching, lazy-loading
DevOps ci-cd, docker, kubernetes
Documentation api-docs, readme, changelog

Commands (94)

Slash commands for quick actions:

Command Purpose
/plan Create implementation plan
/review Code review workflow
/test Test generation
/refactor Refactoring suggestions
/security-scan Security analysis
/cost-report Token usage report

Agents (67)

Specialized AI personas for different tasks:

Agent Purpose
code-reviewer Thorough code review
security-auditor Security analysis
performance-expert Optimization
architect System design
documenter Documentation

Rules (121)

Language-specific coding rules:

Language Rules
Common Best practices, error handling
TypeScript Types, async/await, modules
Python PEP 8, type hints, async
Go Concurrency, error handling
Rust Ownership, lifetimes, traits
Java Design patterns, streams

Usage

Loading Skills

# Load a skill
legion run "use tdd skill to implement feature"

# Or via skill tool
skill(name="tdd")

Using Commands

# Use a command
/plan implement user authentication

# Review code
/review src/auth.ts

Agent Routing

Legion automatically routes tasks to appropriate agents:

User: "Review my code for security issues"
→ Routes to: security-auditor agent
→ Uses: security-scan skill
→ Applies: security rules

Skill Structure

Each skill follows this structure:

---
name: skill-name
description: What this skill does
---

# Skill Title

## When to Use
- Use case 1
- Use case 2

## Instructions
1. Step 1
2. Step 2

## Examples
...

Integration

ELC integrates with Legion’s core systems:

  1. System Prompts - Rules are injected into context
  2. Tool System - Skills provide tool instructions
  3. Agent System - Agents use skills and rules
  4. Session System - Commands are available in sessions

Configuration

Enable/disable ELC in legion.json:

{
  "elc": {
    "enabled": true,
    "skills": true,
    "commands": true,
    "agents": true,
    "rules": true
  }
}

Customization

Adding Custom Skills

Create .legion/skills/my-skill/SKILL.md:

---
name: my-custom-skill
description: Does something specific
---

# My Custom Skill

## Instructions
1. Do this
2. Then that

Disabling Specific Skills

{
  "elc": {
    "disabledSkills": ["legacy-pattern", "deprecated-api"]
  }
}

Examples

TDD Workflow

legion run "use tdd to implement user registration"
# 1. Writes failing test
# 2. Implements minimal code
# 3. Refactors
# 4. Repeats

Security Audit

legion run "security scan this project"
# 1. Scans dependencies
# 2. Checks for vulnerabilities
# 3. Reviews code patterns
# 4. Generates report

Code Review

/review src/auth.ts
# 1. Analyzes code quality
# 2. Checks security
# 3. Reviews performance
# 4. Suggests improvements
This site is open source. Improve this page.