admin

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). Yes — the skill explicitly ingests public, user-generated content as part of its workflows (for example, using npx/npm to install MCP servers and a curl command that reads raw GitHub SKILL.md in references/skills-registry.md), which the agent is instructed to read/interpret and which can materially change tool installation/configuration and subsequent actions.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill includes runtime commands that fetch and execute remote code (e.g., git clone https://github.com/org/mcp-server.git and npx -y @modelcontextprotocol/server-filesystem) which the skill recommends/relies on to install and run MCP servers that execute code and can influence agent behavior, so this is a high-confidence runtime external dependency.