skills/volcengine/openviking/tmux/Gen Agent Trust Hub

tmux

Pass

Audited by Gen Agent Trust Hub on Mar 4, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill enables the execution of arbitrary shell commands within tmux panes using the send-keys command.\n- [DATA_EXFILTRATION]: The skill captures terminal output via capture-pane, which may lead to the exposure of sensitive information like environment variables or secrets displayed during the session.\n- [PROMPT_INJECTION]: The skill has an indirect prompt injection surface because it ingests and processes untrusted terminal data without sanitization or boundary markers.\n
  • Ingestion points: Terminal content is read using tmux capture-pane in SKILL.md and scripts/wait-for-text.sh.\n
  • Boundary markers: No delimiters or instructions to ignore embedded commands are used when captured text is provided to the agent.\n
  • Capability inventory: The skill can execute commands through tmux and perform file system operations via its helper scripts.\n
  • Sanitization: No evidence of filtering or escaping is found for data scraped from the terminal panes.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 4, 2026, 06:00 AM