external-gitcode-ascend-vllm-tests-failure-analysis

Pass

Audited by Gen Agent Trust Hub on May 18, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to perform necessary shell operations for developer tasks, including environment setup via sourcing scripts (set_env.sh), checking hardware status with npu-smi info, and executing regression tests using pytest.- [EXTERNAL_DOWNLOADS]: The skill facilitates the download of model weights and testing assets from established services like Hugging Face (including hf-mirror.com) and ModelScope. It also recommends installing specific Python packages such as runai-model-streamer to resolve test dependencies.- [PROMPT_INJECTION]: The skill has an indirect prompt injection surface (Category 8) because its primary workflow involves ingesting and analyzing external test files and execution logs.
  • Ingestion points: The agent is directed to read test files from the vllm/tests/ directory and log output to /tmp/test_output.log.
  • Boundary markers: No specific delimiters are mandated to isolate untrusted content from the system instructions.
  • Capability inventory: The skill enables file system modifications (copying and patching tests) and the execution of the analyzed code via pytest.
  • Sanitization: There are no explicit instructions to sanitize or validate the content of the external test scripts before analysis or execution.
Audit Metadata
Risk Level
SAFE
Analyzed
May 18, 2026, 07:21 AM
Security Audit — agent-trust-hub — external-gitcode-ascend-vllm-tests-failure-analysis