Install verified Codex planning and completion loops with LazyCodex
Add Codex commands, skills, hooks, diagnostics, and sub-agent roles for deep repository initialization, planning, execution, and evidence-backed completion.
npx skills add agentskillexchange/skills --skill install-verified-codex-planning-and-completion-loops-with-lazycodex
Use LazyCodex when Codex needs a repeatable operating loop for a complex codebase rather than ad hoc chat instructions. The operator installs the package, runs the doctor check, invokes `$init-deep` to generate project memory, uses `$ulw-plan` for decision-complete plans, `$start-work` for durable checklist execution, and `$ulw-loop` when the task must continue until evidence verifies completion. The agent can also select installed roles such as explorer, librarian, planner, reviewer, or specialist skills for focused work. This workflow gives Codex concrete commands to initialize, plan, execute, review, and verify work. The boundary is Codex-specific workflow packaging around skills, hooks, roles, diagnostics, and completion checks; it is not a generic coding-agent product listing.
What this skill actually does
Inputs and prerequisites: Node.js, npm or npx, Codex.
Setup notes: Run npx lazycodex-ai install , verify the setup with npx lazycodex-ai doctor , then invoke the installed Codex commands such as $init-deep , $ulw-plan , $start-work , and $ulw-loop from the Codex composer.
Source and verification boundary: use https://github.com/code-yeongyu/lazycodex as the canonical reference before running the workflow; keep commands, API calls, CLI usage, and generated outputs reviewable against that upstream source.
Framework fit: publish this as a Codex workflow only when the operator can invoke the documented toolchain directly, rather than treating the upstream project as a generic product listing.