external-cannbot-model-model-infer-parallel-impl

Pass

Audited by Gen Agent Trust Hub on Jul 7, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill ingests untrusted configuration data used to guide the model adaptation process.\n
  • Ingestion points: The parallel_config parameters in Step 1 of SKILL.md are sourced from user input or the main agent prompt.\n
  • Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands within the configuration values.\n
  • Capability inventory: The skill facilitates the execution of local scripts infer.sh and utils/weight_convert.sh.\n
  • Sanitization: The skill does not implement validation or sanitization of the configuration values before they are used in the implementation workflow.\n- [COMMAND_EXECUTION]: The implementation workflow includes instructions to execute local shell scripts bash infer.sh and bash utils/weight_convert.sh. These commands are standard for validating model changes and processing weights within the Ascend NPU development environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 7, 2026, 02:39 AM
Security Audit — agent-trust-hub — external-cannbot-model-model-infer-parallel-impl