skills/openclaw/skills/tmux/Gen Agent Trust Hub

tmux

Pass

Audited by Gen Agent Trust Hub on Mar 3, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill facilitates the execution of arbitrary shell commands and keystrokes within tmux sessions using tmux send-keys. This is the primary purpose of the skill, enabling automation of TTY-based applications.
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection. It captures terminal output using tmux capture-pane and processes it to detect completion or state changes. Malicious content within that output (e.g., from logs or untrusted files) could influence the agent's behavior.
  • Ingestion points: Terminal history and output are read via tmux capture-pane in the SKILL.md orchestration examples.
  • Boundary markers: There are no markers or specific instructions implemented to help the agent distinguish between process output and its own instructions.
  • Capability inventory: The skill provides full shell access via tmux, allowing for file system operations, package management, and the execution of other agents.
  • Sanitization: There is no evidence of sanitizing or escaping the captured output before it is interpreted by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 3, 2026, 02:44 PM