Skills
Navigate
skills/duolahypercho/codex-router/codex-subagents/Gen Agent Trust Hub

codex-subagents

Pass

Audited by Gen Agent Trust Hub on Sep 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructions direct the agent to run various shell commands from the local codex-router checkout.\n
  • Examples include calls to ./bin/control subagents, ./bin/model-router codex doctor, and node src/catalog.mjs.\n
  • These commands are used to modify local configuration files and rebuild the model catalog as part of the tool's core functionality.\n- [DYNAMIC_EXECUTION]: The skill utilizes node -e with the --input-type=module flag to execute inline JavaScript for diagnostics.\n
  • The capability check command (node --input-type=module -e 'import { checkAgentCapability } ...') imports local modules (./src/agent-check.mjs) to verify model tool-calling performance.\n
  • The subagent-report.mjs script also executes as a standalone Node.js process to analyze local JSON configuration files.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 28, 2026, 02:12 AM