Skills
Navigate
skills/b-open-io/gemskills/codex-agent-setup/Gen Agent Trust Hub

codex-agent-setup

Pass

Audited by Gen Agent Trust Hub on Aug 12, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes bash scripts (setup.sh and generate.sh) that invoke python3 to run scripts located in the project's root directory.
  • Evidence: scripts/setup.sh and scripts/generate.sh calculate a root path (PLUGIN_ROOT) and execute scripts at ${PLUGIN_ROOT}/scripts/codex-agents/install.py and generate.py respectively.
  • [SAFE]: The behavior of the skill is consistent with its stated purpose of installing and maintaining a custom Codex agent.
  • No network exfiltration or data exposure patterns were detected.
  • No obfuscation or prompt injection attempts were found.
  • The resource access pattern (navigating up from the skill directory to reach project-level scripts) is a standard practice for project-integrated tools.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 12, 2026, 02:05 PM
Security Audit — agent-trust-hub — codex-agent-setup