graphics-api-hooking
Pass
Audited by Gen Agent Trust Hub on Feb 25, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches additional documentation and tool lists from the author's official GitHub repository (gmh5225/awesome-game-security).
- [COMMAND_EXECUTION]: Provides code templates for graphics API interception and function hooking (VTable hooks, WinProc hooks) for DirectX, OpenGL, and Vulkan. These are standard techniques for creating overlays or game security research.
- [PROMPT_INJECTION]: Contains an indirect prompt injection surface as it instructs the agent to fetch and process data from a remote Markdown file. -- Ingestion points: The README.md file at https://raw.githubusercontent.com/gmh5225/awesome-game-security/refs/heads/main/README.md. -- Boundary markers: Not present; the skill does not specify delimiters for the remote content. -- Capability inventory: The agent is tasked with summarizing and providing implementation details based on the fetched content. -- Sanitization: No sanitization or validation of the remote content is performed before processing.
Audit Metadata