alicloud-media-live-test

Pass

Audited by Gen Agent Trust Hub on Mar 11, 2026

Risk Level: SAFECOMMAND_EXECUTIONCREDENTIALS_UNSAFEDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a local Python script (tests/common/compile_skill_scripts.py) to validate the compilation of the target skill's source code.
  • [CREDENTIALS_UNSAFE]: The test instructions require the user to configure AliCloud environment variables (ALICLOUD_ACCESS_KEY_ID, ALICLOUD_ACCESS_KEY_SECRET), which are standard for interacting with AliCloud APIs but constitute sensitive data handling.
  • [DATA_EXFILTRATION]: The skill saves raw API request and response data to the local file system in the output/alicloud-media-live-test/ directory. While intended for debugging, this could lead to the exposure of sensitive account metadata or session information in local logs.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by dynamically reading the target skill's SKILL.md file to identify which API to call.
  • Ingestion points: Reading SKILL.md from skills/media/live/alicloud-media-live (File: SKILL.md).
  • Boundary markers: None present; the skill treats the content of the target file as trusted instructions for API selection.
  • Capability inventory: Execution of arbitrary read-only AliCloud API calls (Describe/List/Get).
  • Sanitization: No explicit sanitization or validation of the API name extracted from the target file is mentioned.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 11, 2026, 10:15 AM