Use this pattern when a dashboard sidebar has many actions, commands, reports, or workflows, and users are more likely to search by goal than by exact feature name.
The core idea is simple:
You are the principal engineer responsible for performing a rigorous, evidence-based review of the supplied codebase. The user may provide only a repository URL, repository name, package name, local path, branch, pull request, commit, tag, or archive. Resolve the exact review target, inspect the implementation as a complete system, execute its real workflows, identify concrete problems, and provide a prioritized repair plan. Assume that substantial portions of the codebase may have been produced by coding agents. Individual files may appear polished while cross-module behavior, state ownership, lifecycle handling, error paths, generated artifacts, and tests remain inconsistent or incomplete. Your job is not merely to comment on code quality. Determine:
Small shell snippets for checking local Codex usage from the terminal. Each script is self-contained so you can review exactly what it reads, which endpoint it calls, and what it prints before running it.
Requires Node.js v24+.
bash codex-usage-reset.sh # reset credits only
They overlap in vocabulary, but they are different machines.
Qwik Components Qwik components are resumable component units.
A Qwik component can have:
component$()
useSignal()repo-work-router is a Codex skill for routing repo-scoped work to the right Codex project thread before file writes begin.
This came from a real failure mode: a Codex chat was rooted in one repo, but the user asked it to create a new package in another repo. Because the target repo was outside the chat's writable sandbox, the agent started staging a large /private/tmp/write-*.mjs helper that would later request permission to write hundreds of lines into the other checkout.
That was the wrong shape of solution. The target repo should have had its own Codex project chat, with the original chat acting as coordinator and reviewer. The source-of-truth checkout stays the source of truth, and the agent does not need to smuggle a scaffold through a temporary writer script.
Git History Rewrite is a Codex skill for safely planning or performing local branch history rewrites by change intent. It uses Git Intent Audit evidence, creates backup branches, preserves attribution, and never force-pushes by default.
Companion skills:
Git Intent Audit is a read-only Codex skill for understanding Git changes before committing or rewriting. It audits large dirty worktrees, unclear commit boundaries, oversized commits, mixed-intent commits, and commit message/diff mismatches.
Companion skills:
Auto Git is a Codex skill for turning repo changes into understandable Git history. It auto-detects the current Git topology, detects existing worktrees, claims cooperative run leases, groups dirty changes by change intent, and routes work through local-review checkpoints or coordinated branch/PR/fanout flows.
Optional companion skills:
git-intent-audit: read-only audit for large dirty worktrees, unclear intent, oversized commits, mixed commits, and message/diff mismatch findings.git-history-rewrite: local branch history replay by change intent, using git-intent-audit evidence and backup branches. It never force-pushes by default.This gist bundles five Codex skills:
implement-code-task: lightweight direct coding workflow for simple-to-moderate tasks.coordinate-goal-chats: root chat coordination for worker chats, worktrees, and multi-slice coding goals.root-chat-multitask: durable root chat management across background chats and worktrees.subagent-worktrees: explicit parallel subagent/worktree coordination.run-parallel-scouts: low-reasoning, read-only scout campaigns for broad reviews or audits when you want cheap parallel evidence before the root chat synthesizes and decides what matters.