route
Pass
Audited by Gen Agent Trust Hub on Aug 24, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes external CLI tools (
codex,where/which, and thecodex-companion.mjsplugin script) using Node.js'sspawnandspawnSyncmodules. - These executions are part of the skill's primary functionality for task delegation.
- The implementation includes defensive measures such as closing
stdinto prevent interactive shell hangs and enforcing hard timeouts on all synchronous calls. - [REMOTE_CODE_EXECUTION]: The skill interacts with an external service (OpenAI Codex) to perform tasks.
- This is the intended purpose of the skill. The execution environment for these tasks is scoped to the
codexenvironment (either via the plugin or the standalone CLI). - It uses a structured wrapper (
route.mjs) to ensure these tasks are launched in the background without blocking the main agent session. - [DATA_EXPOSURE]: The skill stores job logs and metadata in the user's home directory under
~/.claude/route-jobs/. - This is standard behavior for job tracking and does not involve exfiltration of sensitive system credentials.
Audit Metadata