mcp-setup

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly asks the agent to collect API keys and GitHub personal access tokens and instructs constructing CLI commands and HTTP headers that embed those secrets verbatim (e.g., -e EXA_API_KEY=, Authorization: Bearer ), which forces the LLM to handle/output secret values directly.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly fetches and runs remote code at runtime (npx -y @upstash/context7-mcp, npx -y exa-mcp-server, npx -y @modelcontextprotocol/server-filesystem), pulls/executes a Docker image (ghcr.io/github/github-mcp-server), and configures HTTP MCP endpoints that the agent will use at runtime (e.g., https://api.githubcopilot.com/mcp/ and custom URLs like https://example.com/mcp), so these external URLs control prompts/behavior or execute code.