skills/mage-space/skills/mage-refine/Gen Agent Trust Hub

mage-refine

Pass

Audited by Gen Agent Trust Hub on Nov 11, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill serves as a functional interface for the Mage.space image refinement service. It correctly utilizes the official Model Context Protocol (MCP) transport (https://mcp.mage.space/mcp) and requires user authentication via the platform's standard connection flow.
  • [SAFE]: All external references and domains (mage.space) are owned by the verified author, mage-space. No suspicious or third-party domains were detected.
  • [SAFE]: The skill provides clear instructions for error handling, specifically directing the agent to respect content filters and not attempt to bypass them.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user data in the form of images and text instructions.
  • Ingestion points: User instructions and image links processed in SKILL.md.
  • Boundary markers: Present; the skill instructs the agent to use specific preset prompts exactly or to use the user's sentence without the surrounding context.
  • Capability inventory: generate (outbound API calls) and create_upload (local file access for uploads).
  • Sanitization: The skill explicitly instructs the agent to respect content_blocked errors, relying on downstream safety filters.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 11, 2026, 07:17 AM