ascendc-operator-performance-optim
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local shell commands to manage the kernel build cycle, including
source ${ASCEND_HOME_PATH}/set_env.sh,bash build.sh, andpip install output/ascend_kernel*.whl. - [DYNAMIC_EXECUTION]: The skill identifies optimization opportunities in C++ code and performs automated modifications, followed by local compilation and installation of the resulting binaries.
- [INDIRECT_PROMPT_INJECTION]: The skill analyzes local repository files which could be manipulated to include malicious instructions for the agent.
- Ingestion points: Reads
design.md,op_host/<op_name>.cpp, andop_kernel/<op_name>.cpp(SKILL.md). - Boundary markers: No delimiters or isolation instructions are employed when processing these external source files.
- Capability inventory: The skill maintains capabilities to modify files and execute shell commands in the local environment.
- Sanitization: Content from ingested source files is not sanitized or validated for safety before processing.
Audit Metadata