skills/muxy-app/muxy/muxy-cli/Gen Agent Trust Hub

muxy-cli

Pass

Audited by Gen Agent Trust Hub on Nov 19, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides comprehensive instructions for using the muxy session create and muxy session send commands. These allow the agent to start new shell sessions and send arbitrary text or key sequences (including control characters like Ctrl+C) to them, facilitating full shell interaction.
  • [DYNAMIC_EXECUTION]: The muxy browser eval command provides the capability to execute arbitrary JavaScript within a browser tab. This allows the agent to run code in the context of any website opened in the built-in browser.
  • [INDIRECT_PROMPT_INJECTION]: The skill exposes the agent to untrusted data through commands that read external content, creating a potential injection surface.
  • Ingestion points: Data enters the context via muxy session read-screen, muxy session history, muxy browser read, and muxy browser snapshot (which retrieves interactive element details).
  • Boundary markers: No specific delimiters or instructions to ignore embedded commands are mentioned for the content read from terminals or browser tabs.
  • Capability inventory: The agent can execute shell commands, manage Git worktrees, modify application configuration via muxy config, and interact with web pages.
  • Sanitization: There are no instructions or built-in mechanisms described for sanitizing or validating the content retrieved from external sources before it is processed by the agent.
  • [EXTERNAL_DOWNLOADS]: The skill mentions the muxy install-skills command. This command facilitates the retrieval and installation of additional agent skills, which represents a potential vector for introducing unvetted code into the agent's environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 19, 2026, 08:13 AM