code-yeongyu

Designer-turned-developer who crafts stunning UI/UX even without design mockups.

Read-only GitHub triage for issues and PRs. It analyzes all open items, writing evidence-backed reports to /tmp/{datetime}/, with every claim requiring a GitHub permalink as proof. This skill never takes any action on GitHub, only generates reports.

An easter egg command providing information about oh-my-opencode. It can be triggered by 'omomomo', 'about', or 'easter egg'.

omo; the best agent harness, formerly known as oh-my-opencode.

Automates browser interactions for web testing, form filling, screenshots, and data extraction. Use when the user needs to navigate websites, interact with web pages, fill forms, take screenshots, test web applications, or extract information from web pages.

Automates browser interactions with persistent page state. Use for navigating websites, filling forms, taking screenshots, extracting web data, testing web apps, or automating browser workflows.

Removes unused code from this project with ultrawork mode, LSP-verified safety, and atomic commits. Triggers: remove dead code, cleanup, remove unused.

Publishes oh-my-opencode to npm using a GitHub Actions workflow. It accepts <patch|minor|major> as an argument and is triggered by publish, release, deploy, or npm publish events.

A robust 16-agent pre-publish release gate that detects changes, performs deep analysis, and conducts holistic review using multiple agents before every npm publish.

Essential for all Git operations like atomic commits, rebase/squash, and history search. Recommended for use with task() to save context. Triggers include 'commit', 'rebase', 'squash', 'who wrote', 'when was X added', and 'find the commit that'.

Adversarial multi-agent planning skill that self-orchestrates 5 hostile category members for ruthless cross-critique debate, distilling only defensible insights. These insights are then mandatorily handed to the `plan` agent for executable plan formalization, ideal for planning requiring maximum rigor and surfacing of.

Compare HEAD with the latest published npm version and list all unpublished changes.

Team Mode security research skill orchestrates 3 vulnerability hunters and 2 PoC engineers to audit a codebase in parallel, prove exploitability, classify root causes, and calibrate severity. It is used for security review, vulnerability research, exploitability audit, and threat model validation.

This skill manages the full Pull Request (PR) lifecycle, from implementation in a Git worktree, atomic commits, PR creation, verification (CI, review, approval) to merge, with auto-cleanup. Use it whenever implementation work needs to land as a PR.

Read-only GitHub triage for issues and PRs. It analyzes all open items and generates evidence-backed reports, never taking any action on GitHub.

This adversarial multi-agent planning skill self-orchestrates 5 hostile category members for ruthless cross-critique debate, distilling defensible insights. It then mandates handing these insights to the `plan` agent for executable plan formalization, ideal for planning requiring maximum rigor.

This 16-agent pre-publish release gate detects all changes since the last npm release, performs deep per-change analysis, holistic review, and overall release synthesis. It must be used before every npm publish.

This skill manages the full PR lifecycle, from implementation in a git worktree to atomic commits, PR creation, and a verification loop until merge, with auto-cleanup. It is used for any implementation work requiring a PR.

Fetches a URL, saves its body to a temporary file, and returns the file path. This allows the agent to process specific sections using tools like rg/jq/awk, quoting only relevant information, rather than dumping the entire page into the conversation.