ascendc-operator-doc-gen
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from local source files, design documents, and test scripts to generate documentation. If these files are compromised or contain adversarial instructions, they could influence the agent's behavior during the extraction or generation process.
- Ingestion points: The skill reads
csrc/register.cpp,csrc/ops.h,csrc/ops/<op_name>/design.md,csrc/ops/<op_name>/op_host/<op_name>.cpp, andtests/test_<op_name>.py. - Boundary markers: No explicit delimiters or instructions are used to prevent the agent from following commands embedded within the source files.
- Capability inventory: The skill has file read access to the project directory and file write access to
ascend-kernel/csrc/ops/<op_name>/README.md. - Sanitization: There is no evidence of filtering, escaping, or validation of the text extracted from the source files before it is processed into the final documentation output.
Audit Metadata