Documentation

AgentC workflow layer for Claude Code. 20 commands, 34 skills.

Installation

/plugin marketplace add https://github.com/zero20nemvp/claude-code.git
    /plugin install agentc@zero20nemvp-claude-code

Quick Start

# The core loop
    /next  # Get task
    /do    # Execute with TDD
    /done  # Verify & complete

Commands

CommandPurpose
/archive Move a project from active to archive
/assess Assess goal quality and autonomy readiness before autonomous execution
/auto Autonomous mode - Claude drives, human executes assigned tasks
/cancel-ralph Cancel active Ralph Wiggum loop
/clean-gone Clean up stale local branches marked as [gone] (deleted on remote but still exist locally)
/commit Git commit with auto-generated message. Use --push to push, --pr to create PR
/commit-push-pr Full workflow - commit changes, push branch, and create pull request
/create Create north star or goal. Usage: /create north-star OR /create goal [ns-id]
/do Execute task with TDD discipline. Usage: /do [task] [--tier skimmed|semi|full] [--local]
/done Record task completion with verification evidence
/focus Set priority override - front-of-mind north star
/hookify Create hooks to prevent unwanted behaviors. Use --list to show existing rules
/hookify-list List all configured hookify rules
/journal Log observations for pattern detection
/next Get your next optimal task - dynamically generated across all north stars/goals
/now What's happening now? Use --full for comprehensive view
/ralph-loop Start Ralph Wiggum loop in current session
/review-pr Comprehensive PR review using specialized agents. Usage: /review-pr [aspects] [parallel]
/skip Skip current task to work on a queued task instead
/timer Check or control the block timer

Skills (34)

AgentC includes 34 embedded skills organized by purpose.

Core Discipline

  • systematic-debugging
  • test-driven-development
  • testing-anti-patterns
  • testing-skills-with-subagents
  • verification-before-completion

Discovery & Planning

  • brainstorming
  • executing-plans
  • feature-writing
  • jtbd-discovery
  • product-discovery
  • story-mapping
  • vertical-slicing
  • writing-plans

Execution

  • receiving-code-review
  • requesting-code-review
  • using-git-worktrees

Frontend

  • frontend-design
  • lofi-wireframes

Ruby/Rails

  • ilx
  • rails-hotwire
  • rails-solid-stack
  • rbs-types

Meta

  • autonomous-loop
  • condition-based-waiting
  • defense-in-depth
  • dispatching-parallel-agents
  • finishing-a-development-branch
  • rebuild
  • root-cause-tracing
  • sharing-skills
  • subagent-driven-development
  • using-agentc
  • writing-rules
  • writing-skills