atb-aclnn-operator-migration
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external ATB and ACLNN interface documentation to generate migration code. This creates a surface for indirect prompt injection where untrusted content in the ingested documents could attempt to influence the code generation or subsequent build steps.
- Ingestion points: External interface documentation and source files within the ATB repository.
- Boundary markers: Not explicitly defined in the skill instructions.
- Capability inventory: File write operations for C++ source/header files and shell command execution for compilation (cmake/make) and testing.
- Sanitization: No specific sanitization or validation of the ingested document content is described.
- [DYNAMIC_EXECUTION]: The skill's primary function involves generating C++ implementation files for ACLNN runners and modifying existing operation files. It subsequently guides the agent to compile and execute this newly generated code to verify functionality and performance. These activities are localized to the user-provided repository path and are consistent with the skill's stated purpose as a migration utility.
Audit Metadata