parallel-worktree-session
Parallel Worktree Session
Overview
One branch. One worktree. One agent. This is the core constraint that makes parallel development work without file conflicts, port collisions, or context pollution between features.
Core principle: Spin up isolated workspaces first, coordinate after.
Phase 1 — Plan: Feature Decomposition
Before creating anything, define what you're parallelizing:
- List features — name each by what it does, not how
- Verify independence — features that modify the same files are NOT good candidates
- Choose execution mode (see decision table below)
- Build a tracking table:
More from arosenkranz/claude-code-config
homelab-helper
Expert guidance for homelab infrastructure, self-hosting, and Raspberry Pi optimization. Use when recommending self-hosted services, configuring Docker services, setting up reverse proxies, integrating Home Assistant, or troubleshooting homelab networking.
17continuous-learning-v2
Instinct-based learning system that observes sessions via hooks, creates atomic instincts with confidence scoring, and evolves them into skills/commands/agents.
6session-log
Document conversation accomplishments in Obsidian vault with frontmatter, code changes, learning notes, and next steps.
6evolve
Cluster related instincts into skills, commands, or agents
6reclaude
Refactor CLAUDE.md files to follow progressive disclosure principles. Use when a CLAUDE.md file is too long or needs organizational improvement.
5find-skills
Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skill should be used when the user is looking for functionality that might exist as an installable skill.
5