skills/halfwhey/tmux-skill/tmux/Gen Agent Trust Hub

tmux

Pass

Audited by Gen Agent Trust Hub on Apr 5, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill facilitates the execution of arbitrary commands in terminal panes through tmux send-keys and the bin/send-tmux script. This is the core functionality intended for managing shell environments.
  • [DATA_EXFILTRATION]: The skill includes functionality to capture and read the output of tmux panes using bin/read-tmux. While this grants the agent access to potentially sensitive information displayed in a terminal, no evidence was found of instructions to exfiltrate this data to external systems.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it reads untrusted content from terminal panes.
  • Ingestion points: Pane output read via bin/read-tmux.
  • Boundary markers: Absent; output is passed to the agent without explicit delimiters.
  • Capability inventory: Sending keys (bin/send-tmux), managing panes, and session control.
  • Sanitization: Output is capped but not sanitized for instruction-like content.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 5, 2026, 06:53 PM