web-framework-angular-standalone

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references a configuration schema at https://raw.githubusercontent.com/agents-inc/cli/main/src/schemas/metadata.schema.json. This is a vendor-owned resource hosted on GitHub and is used for metadata validation.
  • [DATA_EXFILTRATION]: The code examples include usage of fetch and HttpClient to communicate with local API endpoints (e.g., /api/users). These are standard implementation patterns for the framework and do not exfiltrate sensitive data.
  • [INDIRECT_PROMPT_INJECTION]: The skill acts as a reference for code migration and generation, which involves processing user-supplied source code. While this creates a theoretical surface for indirect prompt injection via code comments, the risk is mitigated as the skill does not grant the agent high-privilege capabilities such as arbitrary command execution or file system modification.
  • Ingestion points: User-provided Angular code snippets and project structure (via SKILL.md and related examples).
  • Boundary markers: Standard Markdown headers and code blocks are used to delimit code.
  • Capability inventory: No shell execution, eval(), or sensitive file write capabilities are requested or used.
  • Sanitization: None specified, relying on platform-level LLM guardrails.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 02:48 PM
Security Audit — agent-trust-hub — web-framework-angular-standalone