onul

Plugin onul from marketplace onul-services in etamong-playground/agent-plugins. Commands read from the repo's manifest — 6 days ago.

Skills (38)

• adopt-dependency

  Upgrade an existing project dependency, verify the resolved artifact and release gap, update locks and validate affected behavior.

• adversarial-review

  Adversarially review existing work with N parallel critics, each from a distinct perspective, then re-verify every critique in the main session and accept only the ones that survive — repeated for 2 rounds. Use when the user says "비판적으로 검토", "여러 관점에서 검토", "적대적 리뷰", "red team this", "critique my work

• audience-review

  Draft or review reader-facing guides, documents, PR descriptions, and product copy against the intended audience and the task that audience needs to complete.

• audio-validation

  Validate an audio tool's numerical output, file handling, offline processing and real-device behavior with controlled fixtures and explicit measurement tolerances. Use for DSP regressions, audio app acceptance or release checks.

• bump-selfhosted-app

  Prepare and, when explicitly authorized, roll out a self-hosted application upgrade with verified artifacts, data compatibility and rollback.

• cleanup-orphan-worktrees

  Inspect and remove explicitly approved stale worktrees and local branches while preserving dirty, ambiguous or concurrently owned work.

• codebase-design

  Shared vocabulary for designing deep modules. Use when the user wants to design or improve a module's interface, find deepening opportunities, decide where a seam goes, make code more testable or AI-navigable, or when another skill needs the deep-module vocabulary.

• commit-cleanup

  Reorganize an owned branch into logical commits while preserving its exact final tree and protecting concurrent work.

• compare-spike

  Settle a technology/library adoption question by building competing spike prototypes and comparing measured facts. Use when the user asks "should we adopt X or keep our own", wants to compare libraries/SDKs/approaches hands-on, or asks for an A/B prototype evaluation (비교 스파이크).

• console-ux

  Review or improve admin consoles and tabular report pages, including filters, pagination, status chips, editable values, forms, copy, representative sample data, and screenshot feedback. Use for management-console UI changes, generated HTML reports with tables, or screenshot-based console feedback.

• convention-review

  Review a change against applicable written repository, organization and agent conventions with rule-cited findings.

• curate-links

  Find and verify a focused reference list, checking relevance, access and useful figures with supported browsing tools.

• diagnosing-bugs

  Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.

• domain-modeling

  Build and sharpen a project's domain model. Use when the user wants to pin down domain terminology or a ubiquitous language, record an architectural decision, or when another skill needs to maintain the domain model.

• explain-code-change

  Create a self-contained interactive HTML explainer for a diff, pull request, branch, or commit range. Use when the user explicitly asks for an interactive or offline code walkthrough or onboarding artifact rather than an ordinary chat explanation or code review.

• frontend-rum-privacy

  Audit frontend monitoring and session replay for displayed secrets and personal data, then propose environment-specific controls.

• github-permalink

  Generate verified repository file links with suitable refs, UTF-8 path encoding and line anchors.

• grilling

  Interview the user relentlessly about a plan or design. Use when the user wants to stress-test a plan before building, or uses any 'grill' trigger phrases.

• incident-triage

  Investigate service incidents using authorized logs, metrics and traces, separating observations from hypotheses before proposing recovery.

• korean-prose

  Write and edit natural Korean technical prose, guides, planning documents, PR descriptions, and UI copy while preserving meaning and the requested register.

• macos-app-development

  Plan, implement or verify a macOS app from the user's tasks and explicit requirements, choosing platform integration, UI and tests to fit the project. Use for general Mac app development; add domain workflows only after their features are confirmed.

• macos-audio-app

  Implement explicitly requested macOS audio processing, capture/playback or Audio Unit features with defined file, device and real-time boundaries. Use after audio requirements are confirmed; a sound-engineering role alone does not select this workflow.

• macos-ui-review

  Review or improve native macOS app workflows, keyboard/menu behavior, accessibility and task-state feedback. Use for SwiftUI/AppKit screens; browser screenshots alone cannot validate native controls.

• memory-maintenance

  Audit and reduce agent context and project-note overhead while preserving durable decisions, permissions, references, and resumable work; use for agent memory cleanup, oversized indexes, or stale project notes.

• nextjs-playwright-e2e

  Add deterministic browser tests to a Next.js and Auth.js app with isolated synthetic authentication and representative backend behavior.

• observability-onboarding

  Define and verify a service observability contract using its existing telemetry stack, with bounded labels, correlation and privacy.

• optimize-memory

  Diagnose and reduce a program's memory use with controlled profiling and before/after measurements; use for heap growth, high RSS, allocation pressure, leaks, or excessive GPU memory, not agent note cleanup.

• planning

  Plan a feature, fix, or implementation task from repository evidence, with scope, acceptance criteria, dependencies, and a resumable checkpoint; reuse an accepted plan when available.

• platform-plan-doc

  Write or refresh a standing service plan or design document set with qualified terms, implementation status, architecture diagrams, flow maps, functional specifications, and screen specifications. Use for planning-document and design-document requests, including 기획서, 기능설계서, 화면설계서, and 플로우맵; not for

• prototype

  Build a throwaway prototype to flesh out a design — a runnable terminal app for state/business-logic questions, or several radically different UI variations toggleable from one route.

• prune-comments

  Audit the comments in the regions a branch changed, or in a whole file, path or repo on request, and delete the ones the code already says. Use when the user says "불필요한 코멘트 정리/제거", "주석 정리해줘", "prune the comments", "remove redundant comments", "clean up the comments", or when a change has left behind

• refactor-safely

  Refactor existing code without changing observable behavior, using characterization tests and small named transformations. Use when the user explicitly asks to restructure, modernize, or clean up code while preserving what it does; use architecture skills to survey opportunities or choose new bounda

• repo-ci-baseline

  Audit whether pull-request CI actually executes meaningful tests and prepare the appropriate enforcement through the repository workflow.

• requirements-feedback

  Turn user interviews and feedback on an existing feature or specification into traceable requirement changes and acceptance examples. Use when feedback must be reconciled with a plan, prototype or implemented behavior.

• tdd

  Test-driven development. Use when the user wants to build features or fix bugs test-first, mentions "red-green-refactor", or wants integration tests.

• teach

  Teach the user a new skill or concept, within this workspace.

• update-agent-guidance

  Turn an explicit request or recurring correction into a concise, correctly scoped agent rule or reusable skill, then verify the intended clients can load it.

• writing-great-skills

  Reference for writing and editing skills well — the vocabulary and principles that make a skill predictable.

Similar plugins