csp-lua-sdk
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied scripts for debugging and modification, creating an ingestion point for potentially malicious instructions hidden within code comments or logic. 1. Ingestion points: User-provided Lua scripts and prompts within the agent conversation. 2. Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded commands in user scripts. 3. Capability inventory: The skill references APIs for network communication (web.get), file system operations (ac.readFile, ac.scanDir), and dynamic code execution (loadstring). 4. Sanitization: No specific sanitization or validation logic is provided for external script content.
- [DATA_EXFILTRATION]: The SDK documentation in Quick_Reference.md describes APIs for file operations (ac.readFile, ac.scanDir) and web requests (web.get), which are standard for Assetto Corsa development but represent a potential exfiltration path if misused.
- [DYNAMIC_EXECUTION]: The Reference_Manual.md file contains documentation for using the loadstring function, which allows for the execution of code generated at runtime.
Audit Metadata