mcporter

Warn

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill relies on npx --yes mcporter to fetch and execute the MCPorter CLI from the npm registry if it is not already installed. This occurs in the run_mcporter function within scripts/mcporter.sh.
  • [COMMAND_EXECUTION]: The script scripts/mcporter.sh uses a Python inline script to execute shell commands via subprocess.run(). This pattern is used to implement a timeout mechanism for the MCP tool calls.
  • [DATA_EXPOSURE]: The skill's primary purpose is to discover and interact with MCP server configurations from various sources like OpenCode, Cursor, and Claude. This involves reading local configuration files (e.g., ./config/mcporter.json) which may contain sensitive connection details or tool definitions.
  • [INDIRECT_PROMPT_INJECTION]: The skill acts as a bridge to external MCP tools. Since it can call arbitrary tools with user-provided arguments (e.g., firecrawl.scrape url=https://example.com), it provides a surface where untrusted data from the internet could be ingested into the agent's context through the output of those tools.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 13, 2026, 07:25 AM