ruview-rvagent

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). The skill deliberately spawns a third‑party npm package (npx -y @ruvnet/rvagent) that exposes sensitive sensor/vitals surfaces and registers as an MCP server — creating clear supply‑chain, remote‑code‑execution, and data‑exfiltration/backdoor risk vectors.

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

• Third-party content exposure detected (high risk: 0.75). The required runtime path is that Claude Code spawns the published MCP server @ruvnet/rvagent (npx -y @ruvnet/rvagent), which then calls the sensing server at RVAGENT_SENSING_URL (default http://localhost:3000) and ingests BFLD event data (e.g., via bfld_last_scan / bfld_subscribe) into tool outputs that become LLM context; this is outsider-authored free text/data fetched from an external service not authored by the operating user.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill spawns "npx -y @ruvnet/rvagent" at activation, which fetches and executes a remote npm package at runtime (npx/@ruvnet/rvagent), so external code is executed and used as a required dependency.