code-yeongyuCommunity·11 Agent Skills Included

lazycodex

Turn Codex into a full AI dev team with verified completion

Installs a complete agent harness for Codex: project memory, planning, parallel sub-agent execution, and evidence-verified completion. Replaces manual prompting, unverified 'done' claims, and lost context with durable plans, hooks, and proof artifacts. Adds specialist skills for debugging, refactoring, frontend design, data analysis, visual QA, and Git history work.

3,363|211

↗

npx skills add code-yeongyu/lazycodex --all -g -y

Available:

Explains how the session-start bootstrap provisions the agent runtime and how the teammode component routes the agent to coordinate named teams of Codex workers.

All Skills in This Repository (11)

Pure Emerald Level Indicators

📦 In Repo

code-yeongyu

ulw-plan

Converts vague planning requests into a single decision-complete work plan.

teammode

Coordinate Codex threads as a named team with durable on-disk state.

lcx-report-bug

Route LazyCodex or Codex bugs to the correct repository with runtime evidence.

lcx-doctor

Diagnose LazyCodex/Codex installation health against latest upstream sources.

ast-grep

Perform AST-based code search and codemods with offline pattern validation.

frontend

Route frontend tasks through triage gates and load design references.

ultraresearch

Coordinate parallel research agents across codebases, web, docs, and OSS repositories.

visual-qa

Identify visual fidelity gaps between intended design and rendered UI.

lcx-contribute-bug-fix

Reproduce LazyCodex and Codex defects and generate minimal verified patches.

lsp-setup

Configure a Language Server for a project by detecting file extensions and writing per-project configs.

start-work

Orchestrates parallel subagent execution of Prometheus work plans with evidence tracking in Codex.

Community

Advanced

Frequently Asked Questions

FAQPage Schema

How to install LazyCodex?▼

Run `npx skills add code-yeongyu/lazycodex --all -g -y` in your terminal to install the full skill suite globally.

What does LazyCodex add to Codex?▼

It adds project memory via AGENTS.md files, a planning command, durable plan execution, and self-verifying work loops. It also installs agent roles, hooks, and specialist skills so Codex behaves like a coordinated dev team.

How do I run a plan to completion with AI?▼

Use $ulw-plan to produce a decision-complete plan, then $start-work to execute every checkbox with evidence. The $ulw-loop command keeps work moving until completion is verified by real proof, not status claims.

Can LazyCodex run multiple agents in parallel?▼

Yes. The teammode skill coordinates a named team of Codex workers with durable shared state, and sub-agent roles like explorer, librarian, and reviewer can be spawned for parallel work.

Do I need coding experience to use LazyCodex?▼

No. After installation you invoke skills with plain-English requests or simple $ commands, and the agent handles planning, execution, and verification for you.