AgentX

Socratic deep interview with mathematical ambiguity gating before execution

Diagnose and fix oh-my-codex installation issues

Save notes to notepad.md for compaction resilience

Setup and configure oh-my-codex using current CLI behavior

Curate and maintain AgentX interview question docs. Use when the user wants to archive new interview questions, merge them into existing topic docs, create new topic docs when needed, or keep answers aligned with current truth vs target design.

Create or revise AgentX stack capability profiles, SQL companions, scenario pack skeletons, and fixture checklists. Use when adding a new technical-stack capability, migrating an existing stack into the profile system, or updating a profile after evaluation reports expose gaps in runtime, catalog, prompt, or eval wiring.

Create or extend AgentX strict real-workflow eval scenario packs and their supporting fixture/runner wiring. Use when the task is to add a new end-to-end eval scenario under src/test/resources/evaluation/scenarios, update the strict real runner, or define how a new real workflow case should be seeded, stopped, and reported.