NYC

ai-model-web

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [External Downloads] (SAFE): The skill references '@cloudbase/js-sdk'. This is an expected, well-known package for Tencent Cloud services and is used according to its primary purpose.
  • [Data Exposure] (SAFE): No hardcoded secrets or sensitive file paths were found. Credentials are provided as placeholders, and the skill includes documentation advising the use of publishable keys over secret keys.
  • [Indirect Prompt Injection] (SAFE): The skill defines an interface for passing user input to AI models. 1. Ingestion points: messages property in the generateText and streamText examples in SKILL.md. 2. Boundary markers: Absent in boilerplate snippets. 3. Capability inventory: Capability to call remote AI models via the CloudBase SDK. 4. Sanitization: Not present in the documentation. Since this represents the primary intended function of the skill and no malicious instructions are embedded, the risk is categorized as safe.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 04:51 PM