plugin-registry
Claude Code plugin registry structure, installation scopes, and common issues. Use when troubleshooting plugin installation problems or manually fixing registry entries.
What this skill does
# Claude Code Plugin Registry
Expert knowledge for understanding and troubleshooting the Claude Code plugin registry.
## When to Use This Skill
| Use this skill when... | Use something else when... |
|------------------------|---------------------------|
| Plugin shows "installed" but isn't working | Setting up new plugins (use `/configure:claude-plugins`) |
| Need to understand plugin scopes | Configuring plugin permissions (use settings-configuration skill) |
| Fixing orphaned registry entries | Creating workflows with plugins (use github-actions-plugin) |
| Debugging installation failures | |
## Registry Location
The plugin registry is stored at:
```
~/.claude/plugins/installed_plugins.json
```
This file tracks all installed plugins across all projects.
## Registry Structure (v2)
```json
{
"version": 2,
"plugins": {
"plugin-name@marketplace-name": [
{
"scope": "project",
"projectPath": "/path/to/project",
"installPath": "~/.claude/plugins/cache/marketplace/plugin-name/1.0.0",
"version": "1.0.0",
"installedAt": "2024-01-15T10:30:00Z",
"lastUpdated": "2024-01-15T10:30:00Z",
"gitCommitSha": "abc123"
}
]
}
}
```
Each plugin key maps to an **array** of installations (supporting multiple scopes).
### Field Reference
| Field | Required | Description |
|-------|----------|-------------|
| `scope` | Yes | `"project"` or `"user"` (global) |
| `projectPath` | project only | Directory where plugin is active |
| `installPath` | Yes | Cache path for installed plugin files |
| `version` | Yes | Installed version |
| `installedAt` | Yes | ISO timestamp of installation |
| `lastUpdated` | Yes | ISO timestamp of last update |
| `gitCommitSha` | Yes | Git commit of installed version |
## Installation Scopes
### User Scope (global, default)
```bash
/plugin install my-plugin@marketplace
```
- `"scope": "user"` in registry entry
- No `projectPath` field
- Available in all projects
### Project Scope
```bash
/plugin install my-plugin@marketplace --scope project
```
- `"scope": "project"` in registry entry
- Has `projectPath` set to installation directory
- Should only be active in that project
- **Bug #14202**: Still shows as "installed" in other projects
## Known Issue: #14202
**Problem**: Project-scoped plugins incorrectly appear as globally installed.
**Root Cause**: Inconsistent `projectPath` checking:
| Operation | Checks projectPath? | Result |
|-----------|---------------------|--------|
| Marketplaces "(installed)" | No | Shows installed everywhere |
| `/plugin install` | No | Refuses to install |
| Installed tab listing | Yes | Correctly filtered |
**Symptoms**:
1. Plugin shows "(installed)" checkmark in Marketplaces view
2. `/plugin install` says "already installed"
3. Plugin doesn't appear in Installed tab for current project
4. Plugin doesn't actually work in current project
**Workaround**: Manually edit the registry to add an entry for the current project.
## Manual Registry Operations
### View Registry
```bash
jq . ~/.claude/plugins/installed_plugins.json
```
### List All Plugins
```bash
jq -r '.plugins | keys[]' ~/.claude/plugins/installed_plugins.json
```
### Find Project-Scoped Plugins
```bash
jq '.plugins | to_entries[] | .value[] | select(.scope == "project") | {projectPath, version}' ~/.claude/plugins/installed_plugins.json
```
### Find Orphaned Entries
Use the Read tool to read `~/.claude/plugins/installed_plugins.json`, then check each `projectPath` with `test -d`.
### Backup Registry
```bash
cp ~/.claude/plugins/installed_plugins.json ~/.claude/plugins/installed_plugins.json.backup
```
## Fixing Registry Issues
### Remove Orphaned Entry
1. Read `~/.claude/plugins/installed_plugins.json` with the Read tool
2. Back up with `cp ~/.claude/plugins/installed_plugins.json ~/.claude/plugins/installed_plugins.json.backup`
3. Remove the orphaned entry from the `plugins` object
4. Write the updated JSON with the Write tool
### Add Entry for Current Project
1. Read the registry with Read tool
2. Add a new entry to the plugin's array with `scope: "project"` and current `projectPath`
3. Write the updated JSON with Write tool
### Convert Project-Scoped to User (Global)
1. Read the registry with Read tool
2. Change `"scope": "project"` to `"scope": "user"` and remove `projectPath`
3. Write the updated JSON with Write tool
## Project Settings Integration
Project-scoped plugins also need entries in `.claude/settings.json`:
```json
{
"enabledPlugins": [
"plugin-name@marketplace"
]
}
```
Without this, even a correctly registered project-scoped plugin won't load.
## Troubleshooting Checklist
1. **Plugin shows installed but doesn't work**
- Check if `projectPath` matches current directory
- Check `.claude/settings.json` for `enabledPlugins`
- Run `/health:plugins` for diagnosis
2. **Can't install plugin (already installed)**
- Check registry for existing entry
- Check if entry has different `projectPath`
- Use `/health:plugins --fix` or manual edit
3. **Plugin works in one project but not another**
- Likely a project-scoped plugin
- Need separate registry entry per project
- Or convert to global scope
4. **Registry file is corrupted**
- Restore from backup if available
- Or delete and reinstall plugins
- Location: `~/.claude/plugins/installed_plugins.json`
## Agentic Optimizations
| Context | Command |
|---------|---------|
| View registry | `jq -c . ~/.claude/plugins/installed_plugins.json` |
| List plugins | `jq -r '.plugins \| keys[]' ~/.claude/plugins/installed_plugins.json` |
| Check specific | `jq '.plugins."name@market"' ~/.claude/plugins/installed_plugins.json` |
| Project plugins | `jq '.plugins \| to_entries[] \| .value[] \| select(.scope=="project")' ~/.claude/plugins/installed_plugins.json` |
## Quick Reference
### Registry Path
```
~/.claude/plugins/installed_plugins.json
```
### Key Format
```
{plugin-name}@{marketplace-name}
```
### Scope Indicator
- `"scope": "project"` + `projectPath` → Project-scoped
- `"scope": "user"` → Global (user-wide)
### After Editing
Always restart Claude Code for registry changes to take effect.
Related in AI Agents
skill-development
IncludedComprehensive meta-skill for creating, managing, validating, auditing, and distributing Claude Code skills and slash commands (unified in v2.1.3+). Provides skill templates, creation workflows, validation patterns, audit checklists, naming conventions, YAML frontmatter guidance, progressive disclosure examples, and best practices lookup. Use when creating new skills, validating existing skills, auditing skill quality, understanding skill architecture, needing skill templates, learning about YAML frontmatter requirements, progressive disclosure patterns, tool restrictions (allowed-tools), skill composition, skill naming conventions, troubleshooting skill activation issues, creating custom slash commands, configuring command frontmatter, using command arguments ($ARGUMENTS, $1, $2), bash execution in commands, file references in commands, command namespacing, plugin commands, MCP slash commands, Skill tool configuration, or deciding between skills vs slash commands. Delegates to docs-management skill for official documentation.
reprompter
IncludedTransform messy prompts into well-structured, effective prompts — single or multi-agent. Use when: "reprompt", "reprompt this", "clean up this prompt", "structure my prompt", rough text needing XML tags and best practices, "reprompter teams", "repromptception", "run with quality", "smart run", "smart agents", multi-agent tasks, audits, parallel work, anything going to agent teams. Don't use when: simple Q&A, pure chat, immediate execution-only tasks. See "Don't Use When" section for details. Outputs: Structured XML/Markdown prompt, quality score (before/after), optional team brief + per-agent sub-prompts, agent team output files. Success criteria: Single mode quality score ≥ 7/10; Repromptception per-agent prompt quality score 8+/10; all required sections present, actionable and specific.
adaptive-compaction
IncludedAdaptive add-on policy and recovery layer that decides WHEN to compact, prune, snapshot, or fork -- replacing fixed-percent auto-compaction across Claude Code, Codex, and MCP-capable hosts. Trigger on auto-compact timing or damage: "when should I compact", "is it safe to compact now or start a fresh session", "auto-compact fires too early/mid-task", "switching to an unrelated task but the window still has space", "context rot", "answers get worse the longer the session runs", "the agent forgot the plan or my decisions after it summarized", "add a layer on top that manages context without changing the agent", raising autoCompactWindow to give the policy room, or installing/tuning a cross-tool compaction policy or PreCompact hook -- even when "compaction" is never said but the problem is context-window pressure or post-summarization memory loss. Do NOT use to summarize a conversation, build RAG, write a summarization prompt (decides WHEN not HOW), or answer max-context-length trivia.
agent-skill-creator
IncludedCreate cross-platform agent skills from workflow descriptions. Activates when users ask to create an agent, automate a repetitive workflow, create a custom skill, or need advanced agent creation. Triggers on phrases like create agent for, automate workflow, create skill for, every day I have to, daily I need to, turn process into agent, need to automate, create a cross-platform skill, validate this skill, export this skill, migrate this skill. Supports single skills, multi-agent suites, transcript processing, template-based creation, interactive configuration, cross-platform export, and spec validation.
llm-wiki
IncludedUse when building or maintaining a persistent personal knowledge base (second brain) in Obsidian where an LLM incrementally ingests sources, updates entity/concept pages, maintains cross-references, and keeps a synthesis current. Triggers include "second brain", "Obsidian wiki", "personal knowledge management", "ingest this paper/article/book", "build a research wiki", "compound knowledge", "Memex", or whenever the user wants knowledge to accumulate across sessions instead of being re-derived by RAG on every query.
skill-master
IncludedAgent Skills authoring, evaluation, and optimization. Create, edit, validate, benchmark, and improve skills following the agentskills.io specification. Use when designing SKILL.md files, structuring skill folders (references, scripts, assets), ingesting external documentation into skills, running trigger evals, benchmarking skill quality, optimizing descriptions, or performing blind A/B comparisons. Keywords: agentskills.io, SKILL.md, skill authoring, eval, benchmark, trigger optimization.