rocksky

Pass

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill documentation instructs the user to install the @rocksky/cli and @rocksky/mcp packages globally via NPM to facilitate the MCP server connection.
  • [COMMAND_EXECUTION]: The instructions provide specific shell commands to install dependencies, log into the Rocksky service, and register the server with the agent's environment (e.g., claude mcp add).
  • [INDIRECT_PROMPT_INJECTION]:
  • Ingestion points: The skill fetches external data such as track titles, artist names, and account stats from the Rocksky service, including data from other users when the actor parameter is utilized in tools like get_scrobbles or get_now_playing (found in SKILL.md).
  • Boundary markers: There are no specified delimiters or explicit instructions provided to the agent to treat data retrieved from external tools as untrusted content.
  • Capability inventory: The skill allows for significant control over the environment, including music playback (now/next/last), volume adjustment, and equalizer settings.
  • Sanitization: The instructions do not describe any sanitization or validation of the metadata retrieved from the hosted service before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 11, 2026, 05:09 AM
Security Audit — agent-trust-hub — rocksky