skills/ascend-ai-coding/awesome-ascend-skills/external-cannbot-ops-ascendc-api-best-practices/Gen Agent Trust Hub
external-cannbot-ops-ascendc-api-best-practices
Pass
Audited by Gen Agent Trust Hub on Aug 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill functions as a static knowledge base for Ascend C development. It provides guidance on arithmetic operations, buffer management, data movement (DataCopy), and precision conversion. It contains no executable code or automated scripts that interact with the host system.
- [DATA_EXPOSURE]: Analysis of the skill's content revealed no hardcoded credentials, sensitive file paths (such as .ssh or .aws), or unauthorized network communication patterns. External references are limited to legitimate code hosting platforms and internal project documentation.
- [REMOTE_CODE_EXECUTION]: The skill does not attempt to install external packages (npm, pip, etc.) or execute remote scripts (curl | bash). The C++ code snippets included are purely for instructional use and do not contain hidden execution triggers.
- [PROMPT_INJECTION]: No attempts to override system instructions, bypass safety filters, or extract system prompts were found. The instructional language is professional and focused on the technical domain of hardware-accelerated computing.
- [OBFUSCATION]: The skill contains no encoded strings (Base64), zero-width characters, homoglyphs, or other techniques designed to hide malicious content from static analysis.
Audit Metadata