Wraps the ai-scanner CLI tool to expose three distinct operations: scan_directory does a full sweep for LLM SDKs, AI frameworks, exposed tokens, and hardcoded secrets with severity ratings. check_secrets runs a focused pass/fail security check that's meant for pre-commit hooks. ai_inventory catalogs which AI SDKs, frameworks, models, and API endpoints your codebase uses without looking for secrets. Covers 20+ AI-specific tokens like OpenAI and Anthropic keys, 59 generic secrets including Stripe and GitHub tokens, plus 23 LLM SDKs and 24 AI frameworks. You'd reach for this when auditing a project's AI dependencies, hunting for exposed credentials before a push, or mapping out what LLM infrastructure a codebase actually relies on.
claude mcp add --transport stdio io.github.aakashbhardwaj27-ai-scanner uvx ai-scanner