Manzanita-Research

Chaparral — Development skill for Claude Code

Development community

🔗 The connective tissue between your projects.

How to install Chaparral

This entry records only its repository, not the path inside it, so there is no exact command to give. Open Manzanita-Research/chaparral and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Chaparral does

🔗 The connective tissue between your projects. Manages shared Claude Code skills and configuration across sibling repos.

Alternatives in Development

Release & Shipping Burn Patterns — Research Summary — Research completed: 2026-02-10 Sources: ~/.claude/CLAUDE.md (This Has Burned You Before table), SESSION_HANDOF 216 ★
A B Copy — Use when you want a marketing-copy A/B — spawns two writer sibling sessions via CCC that draft the SAME messag 130 ★
Cmux Fresh — Clear cmux's persisted session so the next launch opens with fresh/blank windows 107 ★

README

chaparral

The connective tissue between your projects.

Chaparral manages shared [Claude Code](https://claude.ai/code) skills and brand identity across every repo in an organization. One brand repo holds the truth. Chaparral links it everywhere it needs to go.

Built for people who work across multiple orgs and clients. Multi-org by default.

The problem

You have an org directory — say `~/code/manzanita-research/` — with a dozen repos inside. They share a design language, a brand voice, a set of Claude Code skills that shape how AI works with your code. But Claude Code discovers skills per-project (`.claude/skills/`) or globally (`~/.claude/skills/`). There's no org level.

So you copy things around. You forget. Projects drift. The new repo never gets the frontend design skill. The brand voice doc is three versions behind in half your projects.

Claude Code has a plugin marketplace now, but it solves a different problem — distribution to other people and machines. Plugins copy to a cache, require per-repo configuration, and need manual updates after changes. When you're actively developing shared skills across sibling repos, you need something faster.

The fix

Put a `chaparral.json` manifest in your brand repo. Chaparral finds it, reads it, and symlinks your shared skills and org-level `CLAUDE.md` into every sibling repo. Edit a skill in the brand repo and it's instantly live everywhere. No reinstalls, no cache invalidation, no per-repo setup.

~/code/manzanita-research/
├── CLAUDE.md                    ← symlink, managed by chaparral
├── brand/
│   ├── chaparral.json           ← manifest
│   ├── org/
│   │   ├── CLAUDE.md            ← org-wide Claude instructions
│   │   └── skills/
│   │       ├── frontend-design/
│   │       └── brand-voice/
│   └── ...
├── toyon/
│   └── .claude/skills/
│       ├── frontend-design/     ← symlink
│       └── brand-voice/         ← symlink
├── ceanothus/
│   └── .claude/skills/
│       ├── frontend-design/     ← sym