Total Skills
48
Author Profile: dEitY719
Skills published by dEitY719 with real stars/downloads and source-aware metadata.
Total Stars
0
Total Downloads
0
Skills Performance
Comparison chart based on real stars and downloads signals from source data.
agents-md-refactor
0
Stars
0
Downloads
0
devx-rate-limit-guard
0
Stars
0
Downloads
0
devx-resume-after-limit
0
Stars
0
Downloads
0
devx-schedule
0
Stars
0
Downloads
0
devx-visualize
0
Stars
0
Downloads
0
gh-add-ai-metrics
0
Stars
0
Downloads
0
gh-commit
0
Stars
0
Downloads
0
gh-issue-create
0
Stars
0
Downloads
0
Published Skills
agents-md-refactor
Refactor an existing AGENTS.md that is too long, has too much inline content, or lacks nested structure. Use when the user says "my AGENTS.md is too long", "split my AGENTS.md", "optimize my context file", "refactor AGENTS.md", or "/agents-md:refactor". Analyzes the file, proposes a split plan, and executes it — creating nested AGENTS.md files and slimming the root file. Distinct from agents-md:check (audit only) and agents-md:create (greenfield).
Repository SourceNeeds Review
devx-rate-limit-guard
[Claude Code Only] Wrap a long-running command with a rate-limit safety net. Schedules a `CronCreate(durable=true)` that fires at the user-supplied reset time, runs the wrapped command, and clears the safety net on successful completion. Use when the user runs /devx:rate-limit-guard, /devx-rate-limit-guard, or asks "rate limit 걸려도 자동 재개", "퇴근하면서 작업 시키고 limit 풀리면 이어가게 해줘", "auto-resume after my token limit resets". Supports multi-cycle re-arming via `--max-cycles N`. Requires the user to know their reset time (run /usage first). Accepts `-h`/`--help`/`help` to print usage.
Repository SourceNeeds Review
devx-resume-after-limit
[Claude Code Only] Companion to /devx:rate-limit-guard. Invoked by the scheduled cron when token-limit reset arrives — verifies the worktree/ branch context recorded by the guard, pre-emptively re-arms the next cron for multi-cycle runs (`--max-cycles N` > 1), re-runs the original wrapped command idempotently, and clears state on success. Use when the user runs /devx:resume-after-limit, /devx-resume-after-limit, or when a /devx:rate-limit-guard cron prompt fires. Accepts an optional <command> argument (cron path) or reads `.claude/.rate-limit-guard.json` (manual re-trigger). Accepts `-h`/`--help`/`help` to print usage.
Repository SourceNeeds Review
devx-schedule
[Claude Code Only] Schedule a skill or command to run after a specified delay (default: 5 min). Requires the `CronCreate` tool — does not work on Codex / Gemini CLIs. Use when the user says "N분 후에 /skill 실행해", "M분 뒤에 [skill] 수행해", "/devx:schedule", "/devx-schedule", "schedule /skill in N minutes", or "[command] N분 후에 해줘". Always invoke this skill whenever the user wants to defer any slash-command or task to run after a time delay.
Repository SourceNeeds Review
devx-visualize
Create beautiful, self-contained HTML visualizations from any content or idea. Use for slide decks, presentations, infographics, dashboards, flowcharts, diagrams, timelines, comparison tables, data visualizations, landing pages, one-pagers, org charts, mind maps, process flows, kanban boards, report summaries, or any visual that helps humans digest information faster. Trigger on requests like "/devx:visualize", "visualize this," "make a deck," "create a slide," "build an infographic," "show me a dashboard," "make this visual," or any request to present information in a visual HTML format.
Repository SourceNeeds Review
gh-add-ai-metrics
Backfill the ai-metrics footer (tokens · human-h · ai-min) into pre-existing GitHub Issues and PRs that were created before issue #317 / PR #320 introduced automatic metric capture. Use when the user runs /gh:add-ai-metrics, /gh-add-ai-metrics, or asks "기존 이슈에 메트릭 소급 부착", "지난주 PR 들에 ai-metrics 붙여줘", "retrofit metrics on issue #N", "backfill ai-metrics". Three call patterns: (1) no args → infer targets from current chat context; (2) explicit list `issue#N pr#M` (case-insensitive `issue#`/`pr#` prefix); (3) `--type [issue|PR] --date <YYYY-MM-DD>` for date-bounded bulk enumeration via `gh {issue,pr} list --search`. Idempotent — skips cards that already carry an `<!-- ai-metrics -->` block. With `--force`, re-computes metrics fresh and replaces the existing block in place (recompute, not blind overwrite). Never modifies issue/PR body content other than the footer. Accepts `-h`/`--help`/`help` to print usage.
Repository SourceNeeds Review
gh-commit
Create a git commit for the current changes following the repo's style, auto-linking a GitHub issue number if it appears in the recent conversation or is passed as an argument. Always inspects the current working-tree state first — works equally for changes made by Claude in this conversation and for changes the user made manually outside the conversation (e.g. quick alias additions). Use when the user runs /gh:commit, /gh-commit, or asks "커밋해", "지금까지 작업 커밋", "이슈 N번 연결해서 커밋". The user does NOT need to prefix with "git status 확인하고" — that is step 1 of this skill. Creates a new commit — never amends. Never skips hooks. Accepts `-h`/`--help`/`help` to print usage.
Repository SourceNeeds Review
gh-issue-create
Save the current conversation as a GitHub issue in the current repository. Use when the user runs /gh:issue-create, /gh-issue-create, or asks to "이 대화 이슈로 등록", "chat을 깃허브 이슈로 남겨", "기록용 이슈 만들어". Summarizes the conversation so far into a structured issue body keyed by conventional-commit prefix (feat / fix / refactor / perf / docs / test / chore / misc), creates it via `gh issue create` on the target remote's repo without asking for confirmation, and prints only the issue number and URL. Do NOT over- compress — the issue is reused for PR drafts and blog posts, so preserve reasoning, decisions, and concrete details. Accepts an optional remote name argument (e.g., `/gh-issue-create upstream`) to target a different remote's repository instead of origin. When the target repo ships a `.gh-issue-defaults.yml`, default labels and a milestone are auto-applied per that file (Step 2.5); pass `--no-auto-labels` to opt out or `--auto-label-debug` for the dispatch trace. Accepts `-h`/`--help`/`help` to print usage.
Repository SourceNeeds Review
gh-issue-flow
Composition skill that chains gh:issue-implement → gh:commit → gh:pr → devx:schedule (pr-reply, 10 min) → gh:pr-resolve-conflict for a single issue number. Use when the user runs /gh:issue-flow, /gh-issue-flow, or asks "issue #16 처음부터 PR까지 자동으로", "이슈 구현하고 커밋하고 PR까지 한방에", "full flow on #42". Uses direct implementation mode only — for plan/brainstorming modes, use the atomic gh:issue-implement skill manually. Stops on first step failure with a resume-instructions report. Precondition: already on a feature branch in a dedicated worktree. Accepts `<issue-number> [remote]` and `-h`/`--help`/`help`.
Repository SourceNeeds Review
gh-issue-implement
Read a GitHub issue by number and implement it — editing files and running tests, but NOT committing or opening a PR. Use when the user runs /gh:issue-implement, /gh-issue-implement, or asks "issue #16 구현해", "PR 없이 이 이슈 코드만 짜줘", "implement #42". Default mode is direct (no human intervention); optional `plan` or `brainstorming` modes invoke the matching superpowers skills when the plugin is installed (falls back to direct with a warning if not). Precondition: user is already inside a dedicated git worktree on a feature branch. Accepts `<issue-number> [direct|plan|brainstorming] [remote]`, optional `--no-next-hint` (suppress final `Next:` hint), and `-h`/`--help`/`help`.
Repository SourceNeeds Review
gh-issue-read
Fetch a GitHub issue by number and print a structured, human-readable summary without modifying it. Use when the user runs /gh:issue-read, /gh-issue-read, or asks "이슈 #N 읽고 정리해줘", "issue 42 뭐하는 거야", "#16 요약", "이 이슈 내용 파악". Preserves body and comments verbatim so the output can be reused as context for implementation work. Accepts `<issue-number> [remote]`; defaults remote to `origin`. Accepts `-h`/`--help`/`help` to print usage.
Repository SourceNeeds Review
gh-pr-approve
Review a GitHub PR, approve it when clean, request changes for blockers, or file follow-up issues for non-blocking concerns. Use for /gh-pr-approve, /gh:pr-approve, "approve PR 99", "#99 리뷰 승인", or "re-review requested". Self-authored PRs cannot be approved; they use analysis-only, comment-only, or admin-merge paths.
Repository SourceNeeds Review
gh-pr-merge-emergency
Emergency-merge a GitHub PR by bypassing branch-protection approval requirements via admin override, while forcing an audit trail: a reason comment on the PR and a follow-up incident issue for later retro. Use when the user runs /gh-pr-merge-emergency, /gh:pr-merge-emergency, or asks "긴급 머지", "주말 핫픽스 머지", "approval 없이 머지", "admin bypass merge". NOT a replacement for normal review — the skill actively blocks overuse by requiring a written reason and creating a post-merge issue. Required CI must still pass; conflicts still stop the merge. Accepts `-h`/`--help`/`help` to print usage. Project-agnostic; works in any repo where the caller has admin/merge permission.
Repository SourceNeeds Review
gh-pr-merge
Merge an approved GitHub PR using one of three strategies — rebase (default), squash, or merge commit — without asking for confirmation. Use when the user runs /gh:pr-merge, /gh-pr-merge, or asks "PR 51 머지해", "rebase merge", "squash merge", "#99 머지". Refuses to merge un-approved PRs (suggests gh:pr-merge-emergency instead), failing CI, draft PRs, or PRs with conflicts. Accepts `<pr-number> [rebase|squash|merge] [remote]`. Accepts `-h`/`--help`/`help`.
Repository SourceNeeds Review
gh-pr-reply
Fetch code review comments on a GitHub PR, evaluate each one, apply valid fixes, and leave an individual reply on every comment (Accepted with what changed, or Declined with reasoning). Use when the user runs /gh:pr-reply, /gh-pr-reply, or asks "PR 리뷰 코멘트 확인하고 수정", "리뷰 답변 달아", "PR 123 코멘트 처리해". Defaults to the PR for the current branch; accepts an explicit PR number as an argument. Every comment MUST get a reply — bot comments (gemini, sourcery, copilot) included. Accepts `-h`/`--help`/`help` to print usage.
Repository SourceNeeds Review
gh-pr-resolve-conflict
Resolve a GitHub PR's "This branch has conflicts that must be resolved" warning using rebase (never a merge commit), then push with `--force-with-lease`. Use when the user runs /gh:pr-resolve-conflict, /gh-pr-resolve-conflict, or asks "PR conflict 해결", "base 변경됐는데 rebase 해줘", "리베이스로 컨플릭트 풀어". Refuses to run on the default branch, refuses plain `--force`, and never auto-guesses conflict content — the user drives each resolution. Accepts `[pr-number] [remote]`; defaults to the PR attached to the current branch. Accepts `-h`/`--help`/`help`.
Repository SourceNeeds Review
gh-pr
Create a GitHub pull request from the current branch, bundling all commits since it diverged from the base branch. Use when the user runs /gh:pr, /gh-pr, or asks "PR 생성", "풀리퀘 만들어", "지금까지 커밋들로 PR 올려". Pushes the branch if needed, drafts a structured PR body covering every commit in the range (not just HEAD), auto-links a related issue when known, and returns only the PR URL. Accepts `-h`/`--help`/`help` to print usage.
Repository SourceNeeds Review
make-confluence
Transform markdown technical documentation into Confluence-formatted guides with structured problem-solution-results format.
Repository SourceNeeds Review
make-jira
Generate comprehensive weekly Jira reports from work logs.
Repository SourceNeeds Review
mcp-builder
Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).
Repository SourceNeeds Review
pptx
Presentation creation, editing, and analysis. When Claude needs to work with presentations (.pptx files) for: (1) Creating new presentations, (2) Modifying or editing content, (3) Working with layouts, (4) Adding comments or speaker notes, or any other presentation tasks
Repository SourceNeeds Review
skill-refactor
Refactor a SKILL.md that is too long or lacks Progressive Disclosure structure. Shrinks the body to under 100 lines by extracting detail into references/ files. Use when the user says "my skill is too long", "refactor my skill", "apply progressive disclosure to my skill", "slim down my SKILL.md", "/skill:refactor", or after /skill:check reports FAIL/WARN. Distinct from skill:check (audit only).
Repository SourceNeeds Review
web-artifacts-builder
Suite of tools for creating elaborate, multi-component claude.ai HTML artifacts using modern frontend web technologies (React, Tailwind CSS, shadcn/ui). Use for complex artifacts requiring state management, routing, or shadcn/ui components - not for simple single-file HTML/JSX artifacts.
Repository SourceNeeds Review
webapp-testing
Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.
Repository SourceNeeds Review
write-blog-dev-learnings
Write entertaining Korean developer blog posts about debugging war stories, production incidents, and technical gotchas. Saves to ~/para/archive/playbook/docs/dev-learnings/{topic}-blog.md. TRIGGER when user mentions writing a blog about a technical lesson, sharing a debugging experience, or documenting a "삽질" story for teammates. Common triggers include "블로그 써줘", "삽질 블로그", "dev-learnings에 글", "blog post about debugging", "이거 블로그로 정리", "동료한테 공유할 글", "오늘 삽질한 거 글로", or any request to turn a painful technical experience into a shareable narrative. Also trigger when the user recounts a debugging story and wants to preserve it. Do NOT trigger for formal RCA documents (use write:rca), API documentation, README files, or non-narrative technical docs.
Repository SourceNeeds Review
write-insight
Archive one reusable insight from the current conversation into `docs/learnings/<slug>.md` as a short Korean note that follows the repo README rules (5 sections, 50-80 lines, and source links to PR/commit/file:line). Use when the user runs `/write:insight` or asks to capture a concrete pattern, debugging path, or review-driven discovery from this chat. Build the note from conversation evidence (PRs, commits, review threads, repro steps) instead of asking the user to restate context. Decline topics that belong in `docs/technic/`, `docs/standards/`, `docs/feature/`, or `claude/skills/`. Do not use this for narrative "삽질" posts (`write-blog-dev-learnings`), formal postmortems (`write:rca`), or JIRA/PR draft summaries (`write:task-history`).
Repository SourceNeeds Review
write-release-note
Use when generating structured release notes from git history between two releases — finding anchor commits, categorizing by conventional-commit prefix, grouping into user-facing themes, and formatting in Korean
Repository SourceNeeds Review
xlsx
Comprehensive spreadsheet creation, editing, and analysis with support for formulas, formatting, data analysis, and visualization. When Claude needs to work with spreadsheets (.xlsx, .xlsm, .csv, .tsv, etc) for: (1) Creating new spreadsheets with formulas and formatting, (2) Reading or analyzing data, (3) Modify existing spreadsheets while preserving formulas, (4) Data analysis and visualization in spreadsheets, or (5) Recalculating formulas
Repository SourceNeeds Review
ai-worktree-teardown
Clean up an AI worktree after work is done — remove worktree, delete branch, sync main. The reverse of ai-worktree:spawn. Run from the MAIN repo, passing the worktree path as an argument. Use when the user says "작업 끝", "워크트리 정리", "teardown", "cleanup worktree", "작업 완료", "워크트리 제거", or any request to clean up after finishing work in a worktree.
Repository SourceNeeds Review
devx-trd-to-issues
Decompose one or more TRD files (with optional companion PRD) into a three-level Epic → Feature → Task plan and, on `--apply`, register the resulting GitHub Milestones + Issues in bulk via `gh`. Use when the user runs /devx:trd-to-issues, /devx-trd-to-issues, or asks "TRD를 마일스톤/이슈로 분해해줘", "PRD/TRD 파일로 일괄 등록", "decompose this TRD into milestones". Default mode is `--dry-run` — only writes a Markdown plan; `--apply` mutates GitHub. Refuses to auto-create missing labels and refuses an unknown remote. Accepts `-h`/`--help`/`help` to print usage.
Repository SourceNeeds Review
devx-ux-guidelines
Apply UX_GUIDELINES.md standards to shell functions and help text. Trigger on "/devx:ux-guidelines" or when refactoring help functions, creating new help commands, or ensuring consistent formatting with semantic UX functions (ux_header, ux_section, ux_bullet, etc).
Repository SourceNeeds Review
jira-create
Create Jira tickets through the jiravis `jira create-ticket` CLI. Use this skill whenever the user runs /jira:create, /jira-create, asks to create a Jira ticket, says "Jira ticket", "Jira create", "Jira 생성", "Jira 티켓 만들어줘", or wants the current conversation turned into a Jira issue. Requires jiravis CLI installed as `jira`.
Repository SourceNeeds Review
jira-read
Read Jira ticket details through the jiravis `jira get-ticket-detail` CLI. Use this skill whenever the user runs /jira:read, /jira-read, asks to read a Jira ticket, provides a key like ABC-123, says "Jira issue detail", "Jira 읽어줘", "Jira 이슈 파악", or wants a Jira ticket summarized for implementation context. Read-only. Requires jiravis CLI installed as `jira`.
Repository SourceNeeds Review
reverse-engineering-analysis
Analyzes a specific feature or component in the current codebase and generates a ready-to-reuse AI implementation prompt. Use this skill whenever the user wants to understand how a particular feature works, extract the key libraries behind it, or get a copy-pasteable prompt to recreate the same feature in another project. Triggers on phrases like "analyze this feature", "how does X work", "extract the graph feature", "reverse engineer the email system", or any request to understand and reproduce a codebase feature elsewhere. Always produces a markdown document in the specified output directory.
Repository SourceNeeds Review
skill-check
Audit a SKILL.md for structure and UX quality — checks line count, progressive disclosure, frontmatter, references usage, output format, help flag pattern, step structure, options docs, verdict output, and next-action hints. Use when the user says "check my skill", "audit my skill", "does this skill follow best practices?", "/skill:check". Reports PASS/WARN/FAIL/N/A per criterion with concrete fixes. Do NOT use for AGENTS.md or CLAUDE.md files — use agents-md:check or claude-md-check instead.
Repository SourceNeeds Review
agents-md-check
Audit an existing AGENTS.md file for compliance with project documentation standards. Use when the user asks to check, validate, review, or audit an AGENTS.md file — including "is my AGENTS.md good?", "check my context file", "does this follow standards?", or "/agents-md:check". Reports pass/fail/warn per criterion with concrete improvement suggestions. Do NOT use for CLAUDE.md orchestrator files — use claude-md:check instead.
Repository SourceNeeds Review
agents-md-create
Create a new AGENTS.md documentation system for a project from scratch. Use when the user says "create AGENTS.md", "set up my context file", "initialize project docs for AI", "I don't have an AGENTS.md yet", or "/agents-md:create". Generates root AGENTS.md and nested files as needed. Distinct from agents-md:refactor (existing file) and agents-md:check (audit only).
Repository SourceNeeds Review
ai-worktree-spawn
Create an isolated git worktree workspace for the current AI coding agent. Use when starting new parallel work in a multi-AI agent environment. Triggers on: "새로운 작업 시작", "새 작업 시작하자", "격리된 작업 공간 만들어줘", "start new task", "spawn a worktree", or any request to begin isolated work in a multi-agent setup.
Repository SourceNeeds Review
claude-md-create
Create a CLAUDE.md orchestrator file for an AI agent framework from scratch. Use when building a new AI agent orchestration system — multi-agent pipelines, domain-specific orchestrators, automated workflow systems — and you need a well-structured CLAUDE.md to define roles, commands, permissions, and delegation patterns. Triggers on: "create CLAUDE.md", "set up my AI agent framework", "initialize my orchestrator", "build an AI agent system", "/claude-md-create". Distinct from claude-md-check (audit only).
Repository SourceNeeds Review
devx-excalidraw-diagram
Create Excalidraw diagram JSON files that make visual arguments. Trigger on "/devx:excalidraw-diagram" or when the user wants to visualize workflows, architectures, or concepts.
Repository SourceNeeds Review
dissect-builtin
Analyze and document Claude Code built-in skills. Load a built-in skill's prompt via the Skill tool, explain its behavior in Korean, and save structured documentation (README.md + PROMPT.md) to the dotfiles repository. Use when the user wants to study, dissect, or document a built-in skill (e.g., "/dissect-builtin simplify", "/dissect-builtin loop"). Trigger on requests like "내장 스킬 분석", "built-in skill 공부", or "스킬 해부".
Repository SourceNeeds Review
internal-comms
A set of resources to help me write all kinds of internal communications, using the formats that my company likes to use. Claude should use this skill whenever asked to write some sort of internal communications (status reports, leadership updates, 3P updates, company newsletters, FAQs, incident reports, project updates, etc.).
Repository SourceNeeds Review
write-task-history
Write task history from current conversation to a daily task list file. Generates two copy-paste-ready formats: JIRA ticket (plain text with section symbols) and git PR description (markdown). Use this skill whenever the user wants to record, document, or summarize completed work from the current session. Also trigger when the user mentions task history, work log, JIRA ticket drafting from conversation context, or preparing PR descriptions based on what was just done. Works across any project.
Repository SourceNeeds Review
write-rca
Auto-document incidents, bug fixes, and technical challenges as structured markdown with root cause analysis, prevention checklists, and learning resources. Produces Jekyll-compatible publication-ready markdown (YAML frontmatter + single .md files) for postmortem review, technical blogging, AI tool training, and junior engineer onboarding. Saves to ${RCA_REPO_PATH:-~/para/archive/rca-knowledge} with centralized media in _assets/.
Repository SourceNeeds Review
project-setup
Initialize Python projects with standard configuration files (.markdownlint.json, tox.ini, pyproject.toml). Use when starting new Python projects or standardizing existing ones.
Repository SourceNeeds Review
dev-sh-generator
Generate project-specific tools/dev.sh task runner following AGENTS.md standards. Use when creating or updating developer workflow automation scripts for Python/FastAPI projects.
Repository SourceNeeds Review
cli-dev
CLI feature development skill. Implements CLI commands that wrap backend API endpoints. Follows TDD workflow with pytest, Rich console formatting, and session management. Use when implementing REQ-CLI-* requirements.
Repository SourceNeeds Review
sh-check
Audit a shell script (`*.sh`) against the dotfiles quality bar derived from `git_worktree.sh` — the canonical reference implementation. Reports PASS/WARN/FAIL/N/A across 10 criteria covering POSIX hygiene, sourcing guards, naming, ZSH compatibility, help UX, UX-lib usage, input validation, verdict structure, and next-action hints. Use when the user says "check this shell script", "is my .sh file production-ready?", "audit this shell function", "/sh:check". Mirrors `skill:check` but for `.sh` files instead of `SKILL.md`. Do NOT use for SKILL.md (use `skill:check`) or AGENTS.md (use `agents-md:check`).
Repository SourceNeeds Review