My Claude Code plugins, MCP servers, custom HUD statusline, and settings configuration.
git clone https://github.com/yang1997434/Claude-code.git
cd Claude-code
bash install.shOr manually install each component (see below).
A lightweight statusline displayed at the bottom of Claude Code, showing real-time session info at a glance.
# Add this repo as a marketplace
claude /plugin marketplace add https://github.com/yang1997434/Claude-code
# Install HUD plugin — auto-configures everything on first session
claude /plugin install claude-hud@Claude-code4-line layout with ●○ dot progress bars, RGB true color, and vertical | alignment:
Opus 4.6 72k / 200k | 36% used 72,000 | 64% remain 128,000
current: ●●●○○○○○ 32% | weekly: ●○○○○○○○ 16% | sonnet: ○○○○○○○○ 5%
resets 6pm | resets mar 7, 3pm | resets mar 7, 4pm
thinking: Off | cost: $11.05 ↑120k ↓85k | 🔧 12/29 plugins · 1 MCP
| Line | Content | Description |
|---|---|---|
| Line 1 | Opus 4.6 72k / 200k | 36% used 72,000 | 64% remain 128,000 |
Model (color-coded) + token ratio + used% + remain% |
| Line 2 | current: ●●●○○○○○ 32% | weekly: ●○○○○○○○ 16% | sonnet: ○○○○○○○○ 5% |
Subscription quota bars — ●○ dot bars with usage percentage |
| Line 3 | resets 6pm | resets mar 7, 3pm | resets mar 7, 4pm |
Reset times in local time, column-aligned under quotas |
| Line 4 | thinking: Off | cost: $11.05 ↑120k ↓85k | 🔧 12/29 plugins · 1 MCP |
Thinking mode + session cost + plugin/MCP counts |
All lines share column widths — | separators are vertically aligned across all 4 lines.
●○ dot bars and percentages change color (RGB true color) based on usage level:
| Usage | Color |
|---|---|
| < 50% | Green |
| 50–74% | Yellow |
| 75–89% | Orange |
| ≥ 90% | Red (+ ⚠ warning for context) |
- Context, cost, tokens — read from Claude Code's statusLine stdin JSON
- Subscription quotas (current/weekly/sonnet/opus) — fetched from Anthropic OAuth API (
/api/oauth/usage), cached locally for 60s - Reset times — displayed as local time (e.g.
6pm,mar 7, 3:30pm) - Thinking status — read from
~/.claude/settings.json(alwaysThinkingEnabled) - Plugin/MCP counts — read from
~/.claude/settings.jsonand~/.claude/plugins/installed_plugins.json - Credentials —
CLAUDE_CODE_OAUTH_TOKENenv (priority) → macOS Keychain → credentials file, with auto token refresh
# Copy HUD script
mkdir -p ~/.claude/hud
cp hud/custom-hud.mjs ~/.claude/hud/
# Add statusLine to settings.json (or merge manually)
# Ensure settings.json contains:
# "statusLine": {
# "type": "command",
# "command": "node ~/.claude/hud/custom-hud.mjs"
# }Restart Claude Code to activate.
| Plugin | Marketplace | Description |
|---|---|---|
document-skills |
anthropic-agent-skills |
Create/edit docs, slides (PPTX), spreadsheets (XLSX), PDFs, algorithmic art, MCP builders |
superpowers |
claude-plugins-official |
Enhanced agent capabilities, custom agents, hooks |
frontend-design |
claude-plugins-official |
Production-grade frontend UI design and web components |
planning-with-files |
planning-with-files |
Manus-style file-based planning with task_plan.md, findings.md, progress.md |
claude-mem |
thedotmack |
Cross-session persistent memory with semantic search (MCP-based) |
| Plugin | Marketplace | Description |
|---|---|---|
code-review |
claude-plugins-official |
Automated PR review — 5 parallel Sonnet agents with confidence scoring |
pr-review-toolkit |
claude-plugins-official |
Multi-aspect review: code, tests, errors, types, comments, simplify (Opus reviewer) |
review-loop |
hamel-review |
Codex cross-review — independent AI review on task completion |
Install:
claude /plugin install code-review@claude-plugins-official
claude /plugin install pr-review-toolkit@claude-plugins-official
claude /plugin marketplace add https://github.com/hamelsmu/claude-review-loop
claude /plugin install review-loop@hamel-review| Plugin | Language |
|---|---|
pyright |
Python type checking |
pyright-lsp |
Python language server |
vtsls |
TypeScript/JavaScript |
yaml-language-server |
YAML |
Install:
claude /plugin install pyright@claude-code-lsps
claude /plugin install vtsls@claude-code-lsps
claude /plugin install yaml-language-server@claude-code-lsps
claude /plugin install pyright-lsp@claude-plugins-officialOptional LSPs for other languages:
claude /plugin install gopls@claude-code-lsps # Go
claude /plugin install rust-analyzer@claude-code-lsps # Rust
claude /plugin install jdtls@claude-code-lsps # Java
claude /plugin install clangd@claude-code-lsps # C/C++
claude /plugin install omnisharp@claude-code-lsps # C#
claude /plugin install sourcekit-lsp@claude-code-lsps # Swift
claude /plugin install kotlin-lsp@claude-plugins-official # Kotlin
claude /plugin install solargraph@claude-code-lsps # Ruby
claude /plugin install elixir-ls@claude-code-lsps # Elixir
claude /plugin install lua-language-server@claude-code-lsps # Lua
claude /plugin install terraform-ls@claude-code-lsps # Terraform
claude /plugin install gleam@claude-code-lsps # Gleam
claude /plugin install zls@claude-code-lsps # Zig
claude /plugin install dart-analyzer@claude-code-lsps # Dart
claude /plugin install clojure-lsp@claude-code-lsps # Clojure
claude /plugin install bash-language-server@claude-code-lsps # Bash
claude /plugin install intelephense@claude-code-lsps # PHPAll MCP servers are auto-configured by their plugins. No manual .mcp.json editing needed.
| MCP Server | Source Plugin | Tools Provided |
|---|---|---|
mcp-search |
claude-mem |
search, timeline, get_observations, smart_search, smart_unfold, smart_outline |
context7 |
superpowers (external) |
resolve-library-id, query-docs (up-to-date library documentation) |
smart-search |
superpowers |
web_search (multi-engine: Kagi + Tavily) |
Retrieves up-to-date documentation for any library. Usage in Claude Code:
# Automatically available - just ask about any library
"How do I use React hooks?" -> context7 fetches latest React docs
Cross-session memory with semantic search:
# Search past work
/mem-search "how did we implement auth last time?"
# Smart code exploration
/smart-explore "find authentication functions"
Queries Kagi FastGPT + Teclis + TinyGem + Tavily simultaneously. Requires Kagi API key configuration in superpowers plugin settings.
Automated code review gates using Claude Code hooks:
| Hook | Event | Trigger | Action |
|---|---|---|---|
pre-commit-review.sh |
PreToolUse (Bash) | git commit |
Blocks commit unless code review was run (30-min marker) |
post-pr-review.sh |
PostToolUse (Bash) | gh pr create |
Auto-triggers /code-review on the new PR |
Flow:
Write code → git commit → BLOCKED (no review)
↓
code-reviewer (Opus) → fix issues → mark passed → git commit OK
↓
gh pr create
↓
/code-review (5x Sonnet)
Manual install:
mkdir -p ~/.claude/hooks
cp hooks/pre-commit-review.sh ~/.claude/hooks/
cp hooks/post-pr-review.sh ~/.claude/hooks/
chmod +x ~/.claude/hooks/*.shCopy to ~/.claude/settings.json (includes hooks, HUD, and all plugins):
cp settings.json ~/.claude/settings.jsonKey settings:
hooks: Auto-review gates (PreToolUse blocks commit, PostToolUse triggers PR review)statusLine: Custom HUD command — runscustom-hud.mjsto render the statuslineenabledPlugins: Plugins that are active in every sessionCLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS: Enables multi-agent team coordination
| Marketplace | GitHub Repo | Content |
|---|---|---|
anthropic-agent-skills |
anthropics/skills | Document skills (PPTX, XLSX, PDF, etc.) |
claude-plugins-official |
anthropics/claude-plugins-official | Superpowers, frontend-design, external MCP integrations |
planning-with-files |
OthmanAdi/planning-with-files | File-based task planning |
thedotmack |
thedotmack/claude-mem | Persistent cross-session memory |
claude-code-lsps |
boostvolt/claude-code-lsps | Language server plugins (20+ languages) |
hamel-review |
hamelsmu/claude-review-loop | Codex cross-review loop |
Add all marketplaces:
claude /plugin marketplace add https://github.com/anthropics/skills
claude /plugin marketplace add https://github.com/anthropics/claude-plugins-official
claude /plugin marketplace add https://github.com/OthmanAdi/planning-with-files
claude /plugin marketplace add https://github.com/thedotmack/claude-mem
claude /plugin marketplace add https://github.com/boostvolt/claude-code-lsps
claude /plugin marketplace add https://github.com/hamelsmu/claude-review-loopAfter installation, the following skills (slash commands) are available:
| Skill | Plugin | Description |
|---|---|---|
/planning-with-files |
planning-with-files | Start file-based task planning |
/planning-with-files:status |
planning-with-files | Show planning progress |
/mem-search |
claude-mem | Search persistent memory |
/smart-explore |
claude-mem | Token-optimized code search |
/make-plan |
claude-mem | Create implementation plans |
/do |
claude-mem | Execute a phased plan |
/simplify |
superpowers | Review and simplify code |
/code-review |
code-review | Automated PR review with 5 parallel Sonnet agents |
/review-pr |
pr-review-toolkit | Multi-aspect review (code/tests/errors/types/simplify) |
/review-loop |
review-loop | Codex cross-review loop on task completion |
/cancel-review |
review-loop | Cancel an active review loop |
/frontend-design |
frontend-design | Create production-grade UI |
/pptx |
document-skills | Create/edit presentations |
/xlsx |
document-skills | Create/edit spreadsheets |
/pdf |
document-skills | Create/manipulate PDFs |
/docx |
document-skills | Create/edit documents |
/canvas-design |
document-skills | Create visual designs |
/skill-creator |
document-skills | Create new custom skills |
/mcp-builder |
document-skills | Build MCP servers |
/webapp-testing |
document-skills | Test web apps with Playwright |
- Claude Code (latest version)
- Node.js >= 18
- Python >= 3.10 (for pyright LSP)
- Active Anthropic API key or Claude Max subscription