Newsletter
Join the Community
Subscribe to our newsletter for the latest news and updates
Hunt for AI coding artifacts containing secrets by scanning public GitHub repositories for leaked credentials in AI coding tool configuration files.

Autonomous AI agent system for complex penetration testing, integrating security tools, long-term memory, and smart task delegation in sandboxed Docker.

Unified exposure management platform for automated vulnerability scanning, attack surface monitoring, and cloud security posture management.
claudleak is a specialized tool designed to identify and report leaked credentials within AI coding tool configuration files found in public GitHub repositories. It addresses the growing concern of sensitive information exposure through AI-generated or AI-assisted code artifacts.
Key Features:
.claude/, .cursor/, .continue/, .codex/, CLAUDE.md, and AGENTS.md.--verified-only, --org, --user).Use Cases:
How It Works:
Installation:
go install github.com/niels/claudleak/cmd/claudleak@latestOr build from source:
git clone https://github.com/niels/claudleak.git
cd claudleak
go build -o claudleak ./cmd/claudleak/Usage:
GITHUB_TOKEN="ghp_..." ./claudleak