extract-tool-graph

To find hidden tools in a codebase, you need to analyze function handlers, API endpoints, CLI commands, and plugin hooks. This process reveals conditional and dynamic registrations, mapping the full capability surface of your system even when features are not actively running.

Conditional tool registration analysis traces tools activated only under specific conditions like feature flags or environment variables. You need this when developing complex applications to understand the full scope of system capabilities and ensure unexposed features do not introduce security vulnerabilities.

Detect dynamically registered tools by mapping out tools loaded from external sources or dynamically at runtime. This dynamic registration detection requires analyzing plugin hooks and function handlers to reveal capabilities that are not statically defined in your core codebase.

Code analysis can identify security risks by escalating signals when tools with admin-level capabilities are not properly gated. It detects signs of unauthorized data access by mapping the complete capability surface, revealing hidden tools that lack proper security controls.

The best way to map API endpoints and CLI commands for reverse engineering is to extract the complete tool graph. This approach identifies hidden registrations and traces dynamic loading mechanisms, providing a comprehensive view of all system capabilities.

Tool discovery in complex systems is limited by the complexity of tracing dynamic registrations loaded from external sources. Heavily obfuscated runtime loading or deeply nested conditional feature flags may obscure the full capability surface, making complete tool graph extraction challenging.