ggg
Smart context filter - reduce tokens 50-90% before sending to Claude. Use when reading large files (>1000 lines), analyzing complex codebases, or when user types 'ggg'. Saves ~69% cost per task.
npx skills add majiayu000/claude-skill-registry --skill ggg --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
# GGG - Gatekeeper (Smart Context Filter) ## Purpose Reduce token count by 50-90% before sending context to Claude using Gemini Flash or Local LLM. This dramatically reduces costs while maintaining quality. ## When to Use - User explicitly types `ggg` - Before reading large files (>1000 lines) - When analyzing complex codebases - Fixing bugs in specific areas - Adding features to existing code - Any task where you need focused context ## Cost Comparison ``` Before Gatekeeper: - Read 20k tokens: $0.06 (Claude input) - Total: $0.075/task After Gatekeeper: - Filter 20k→2k: $0.0021 (Gemini Flash) - Read 2k tokens: $0.006 (Claude input) - Total: $0.0231/task - SAVINGS: 69% 🎉 ``` ## Steps ### 1. Check Prerequisites Verify Python script exists: ```bash test -f gatekeeper.py && echo "✅ Found" || echo "❌ Missing: gatekeeper.py" ``` ### 2. Get Task Details Ask user (if not already provided): - What file to analyze? - What task to perform? (e.g., "Fix login bug", "Add rate limiting") ### 3. Run Gatekeeper **Mode Selection:** - `auto` (default): Try Gemini Flash → fallback to Local LLM - `flash`: Force Gemini Flash (best quality) - `local`: Force Local LLM (offline/privacy) **Execute:** ```ba
- Purpose
- When to Use
- Cost Comparison
- Steps
- 1. Check Prerequisites
- 2. Get Task Details
- 3. Run Gatekeeper
- 4. Review Output
- 5. Use Filtered Content
- Important Notes
- Quality
- Modes
- Environment Variables
- When NOT to Use
test -f gatekeeper.py && echo "✅ Found" || echo "❌ Missing: gatekeeper.py" python gatekeeper.py [file-path] "[task description]" --mode auto Auto mode (recommended) python gatekeeper.py src/auth.ts "Fix login bug" --mode auto Force Gemini Flash python gatekeeper.py api/routes.ts "Add rate limiting" --mode flash Force Local LLM python gatekeeper.py utils.js "Refactor error handling" --mode local Extract specific keywords python gatekeeper.py large_file.py "token,auth,jwt" --extract
What does the ggg skill do?
Smart context filter - reduce tokens 50-90% before sending to Claude. Use when reading large files (>1000 lines), analyzing complex codebases, or when user types 'ggg'. Saves ~69% cost per task.
How do I install it?
Run `npx skills add majiayu000/claude-skill-registry --skill ggg --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From majiayu000/claude-skill-registry, a repository with 534 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.