skills/opencue/skills/authmux/Gen Agent Trust Hub

authmux

Warn

Audited by Gen Agent Trust Hub on Jun 21, 2026

Risk Level: MEDIUMCREDENTIALS_UNSAFECOMMAND_EXECUTIONDATA_EXFILTRATIONEXTERNAL_DOWNLOADS
Full Analysis
  • [CREDENTIALS_UNSAFE]: The skill's primary function involves accessing, copying, and switching between sensitive authentication tokens located at ~/.codex/auth.json and ~/.codex/accounts/.
  • [DATA_EXFILTRATION]: The authmux export <dir> command allows the agent to copy all stored authentication snapshots to an arbitrary directory, which could be used to move secrets to unsafe locations.
  • [DATA_EXFILTRATION]: The authmux diag command generates a diagnostic bundle. Even if redacted, the collection of system and authentication-related metadata represents an exposure risk.
  • [COMMAND_EXECUTION]: The authmux hook-install command is used to modify shell configuration files (e.g., .bashrc or .zshrc) to persist account-switching logic across terminal sessions.
  • [COMMAND_EXECUTION]: The authmux daemon command initiates a background service that continuously monitors and modifies the active authentication state.
  • [EXTERNAL_DOWNLOADS]: The authmux update command performs network requests to check for and potentially retrieve software updates.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 21, 2026, 12:39 PM
Security Audit — agent-trust-hub — authmux