Skillget

claude-turbo-search

skill
Security Audit
Fail
Health Pass
  • License — License: MIT
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Community trust — 13 GitHub stars
Code Fail
  • rm -rf — Recursive force deletion command in hooks/rag-context-hook.sh
Permissions Pass
  • Permissions — No dangerous permissions requested
Purpose
This tool provides optimized file search and semantic indexing for large codebases directly within Claude Code. It uses utilities like ripgrep and fzf to map projects, enabling faster and more context-aware code retrieval.

Security Assessment
The overall risk is rated as Medium. The tool executes local shell commands to index your project, run dependency checks, and manage files, which carries inherent execution risks. Specifically, the automated scan flagged a `rm -rf` (recursive force deletion) command inside the `hooks/rag-context-hook.sh` script. While this is likely just used for cleaning up temporary files during the indexing process, it warrants a quick manual code review to ensure it doesn't accidentally delete critical project data. No hardcoded secrets were found, and the tool does not request dangerous overarching permissions. It appears to function entirely locally without making suspicious external network requests, though its semantic search dependencies might require standard package downloads.

Quality Assessment
The project is in good health and actively maintained, with its most recent push occurring today. It uses the highly permissive and standard MIT license, making it safe for commercial and personal use. Community trust is currently low but growing, represented by 13 GitHub stars. The documentation is thorough and provides clear, step-by-step instructions for installation and usage across multiple operating systems.

Verdict
Use with caution: the project is active and well-documented, but you should quickly inspect the `rm -rf` command in the shell hook before running it on your machine.
SUMMARY

Optimized file search and semantic indexing for large codebases in Claude Code

README.md

Claude Turbo Search

Optimized file search and semantic indexing for large codebases in Claude Code.

Features

  • Fast file suggestions - ripgrep + fzf for instant autocomplete
  • Semantic search - QMD integration for finding relevant docs by meaning
  • Cartographer integration - Automatic codebase mapping
  • One command setup - /turbo-index does everything
  • QMD skill - /qmd teaches Claude to search before reading files
  • Optional hooks - Auto-inject relevant context before prompts

Requirements

Supported Platforms

Platform Package Manager Status
macOS Homebrew Fully supported
Ubuntu/Debian apt Fully supported
Fedora/RHEL dnf Fully supported
Arch Linux pacman Fully supported
Windows - Not supported (use WSL)

Prerequisites

  • Claude Code CLI installed
  • Bash 4.0+ (default on macOS and Linux)
  • A supported package manager (see above)

Installation

Option 1: Install from GitHub (recommended)

Add the repository as a marketplace and install:

# Add the marketplace from GitHub (use #branch for specific branch)
claude plugin marketplace add iagocavalcante/claude-turbo-search

# Or install from a specific branch
claude plugin marketplace add "iagocavalcante/claude-turbo-search#feature/vector-search-rag"

# Install the plugin
claude plugin install claude-turbo-search@claude-turbo-search-dev

# Restart Claude Code to load the plugin

Option 2: From official marketplace (when published)

claude plugin install claude-turbo-search

Updating the Plugin

When updates are available:

# Update the marketplace to fetch latest changes
claude plugin marketplace update claude-turbo-search-dev

# Update the plugin
claude plugin update claude-turbo-search@claude-turbo-search-dev

# Restart Claude Code to apply updates

Verify Installation

claude plugin list

You should see:

❯ claude-turbo-search@claude-turbo-search-dev
  Version: 1.0.0
  Status: ✔ enabled

Usage

In any project, run:

/turbo-index

This will:

  1. Check and install dependencies (ripgrep, fzf, jq, bun, qmd)
  2. Configure fast file suggestions
  3. Set up QMD MCP server for semantic search
  4. Run cartographer to map the codebase
  5. Index all documentation with QMD

Subsequent runs

Running /turbo-index again will:

  • Skip dependency installation
  • Skip global configuration
  • Refresh the project index if files changed

Available Skills

Skill Description
/turbo-index Set up optimized search indexing for a project
/qmd Search docs before reading to save tokens
/remember Save session context to persistent memory
/memory-stats View memory database statistics
/token-stats Show token economics and savings dashboard
/knowledge-graph Interactive TUI knowledge graph viewer for the memory database

Using the QMD Skill

After indexing, use /qmd or just ask Claude to search:

"Search for authentication logic in this project"
"Find files related to database migrations"

Claude will use QMD to find relevant files before reading them, saving significant tokens.

Using Memory Skills

Track your work across sessions:

# At end of session, save context to memory
/remember

# View accumulated knowledge
/memory-stats

# See token savings in action
/token-stats

The memory system uses SQLite FTS5 for instant search across all your saved sessions, knowledge, and facts.

Knowledge Graph

Visualize entity relationships in your memory database:

# Full dashboard (stats + graph + timeline)
/knowledge-graph

# Individual views
/knowledge-graph stats           # Counts, categories, top entities bar chart
/knowledge-graph graph           # Entity tree + relation edges + co-occurrences
/knowledge-graph timeline        # Chronological session/knowledge entries
/knowledge-graph explore auth    # Drill into a specific entity

Uses Rich for colored TUI output, with automatic plain-text fallback. See docs/knowledge-graph-setup.md for detailed setup.

Manual QMD Commands

# Fast keyword search (use this first)
qmd search "your query" --files -n 10

# Semantic search (slower, use as fallback)
qmd vsearch "how does the login flow work"

# Get specific file content
qmd get "path/to/file.md"

Optional: Auto-Context Hooks

Enable automatic context injection that searches QMD before each prompt:

# Simple mode - suggests relevant file paths (lightweight)
~/claude-turbo-search/scripts/setup-hooks.sh

# RAG mode - injects actual content snippets (recommended)
~/claude-turbo-search/scripts/setup-hooks.sh --rag

# Remove hooks
~/claude-turbo-search/scripts/setup-hooks.sh --remove

Hook Modes Comparison

Mode Token Cost How It Works
Simple ~50-100/prompt Suggests file paths, Claude decides what to read
RAG ~500-2000/prompt Injects content snippets, Claude often needs no file reads

RAG mode is recommended for large codebases - the upfront token cost is offset by avoiding file reads.

How RAG Mode Works

1. You submit: "How does authentication work?"
2. Hook extracts: "authentication work"
3. QMD searches indexed docs
4. Hook injects relevant snippets into context
5. Claude answers using injected context
6. No file reads needed = massive token savings

Dependencies

Tool Purpose
ripgrep Fast file search
fzf Fuzzy finder
jq JSON parsing
bun JavaScript runtime
qmd Semantic search engine

All dependencies are installed automatically on first run using your system's package manager.

How It Saves Tokens

Before (traditional exploration)

Read file1.md (2000 tokens)
Read file2.md (1500 tokens)
Read file3.md (1800 tokens)
→ Found answer in file3.md
Total: 5300 tokens

After (with turbo search)

qmd_search "how does auth work" (50 tokens)
→ Returns: file3.md lines 45-62 (200 tokens)
Total: 250 tokens

Estimated savings: 60-80% on exploration tasks

Sync to a personal web dashboard

Run a single-user dashboard on Fly.io's free tier and have every /remember invocation auto-push your memory.db to it. You then get a browser view of every repo you work in: recent sessions, knowledge areas, facts, and an interactive D3 force-directed knowledge graph.

Deploy on Fly.io

The button takes you to Fly's launch flow. Because the deploy artifact lives at web/, the launch command is:

git clone https://github.com/iagocavalcante/claude-turbo-search.git
cd claude-turbo-search/web
fly launch --copy-config --no-deploy
fly secrets set TURBO_TOKEN=$(openssl rand -hex 32)
fly deploy

Then point the CLI at it once:

~/claude-turbo-search/memory/memory-db.sh config set \
    --remote https://<your-app>.fly.dev \
    --token <your-token>

After this, /remember will silently push to your dashboard whenever it runs. The full architecture, deferred work, and operating notes live in docs/plans/web-sync.md and web/README.md.

Configuration

After running /turbo-index, these files are modified:

  • ~/.claude/settings.json - fileSuggestion and mcpServers config
  • ~/.claude/file-suggestion.sh - turbo file suggestion script
  • .claude/turbo-search.json - project-specific metadata (in each project)

Note: The setup scripts will warn you if existing configuration will be overwritten and create backups automatically.

MCP Tools

After setup, these MCP tools are available:

Tool Description
qmd_search Semantic search across indexed docs
qmd_get Retrieve specific document by path/ID
qmd_collections List all indexed projects

Troubleshooting

Dependencies not installing

If automatic installation fails, you can install dependencies manually:

# macOS
brew install ripgrep fzf jq
brew tap oven-sh/bun && brew install bun
bun install -g https://github.com/tobi/qmd

# Ubuntu/Debian
sudo apt-get install ripgrep fzf jq
curl -fsSL https://bun.sh/install | bash
bun install -g https://github.com/tobi/qmd

# Fedora
sudo dnf install ripgrep fzf jq
curl -fsSL https://bun.sh/install | bash
bun install -g https://github.com/tobi/qmd

QMD models downloading

On first use, QMD downloads ~1.7GB of models. This is normal and only happens once.

Contributing

See CONTRIBUTING.md for guidelines on how to contribute.

License

MIT - see LICENSE for details.

Reviews (0)

No results found