Skip to main content
Manusで任意のスキルを実行
ワンクリックで
GitHub リポジトリ

agent-skills

agent-skills には gbalabanli から収集した 30 個の skills があり、リポジトリ単位の職業カバレッジとサイト内 skill 詳細ページを表示します。

収集済み skills
30
Stars
2
更新
2026-03-19
Forks
0
職業カバレッジ
10 件の職業カテゴリ · 100% 分類済み
リポジトリエクスプローラー

このリポジトリの skills

design-to-code-annotator
ウェブ・デジタルインターフェースデザイナー

Convert annotated UI screenshots into build-ready implementation briefs with screen flow, interaction behavior, and component/state hints for downstream coding.

2026-03-19
google-ai-mode-skill
ソフトウェア開発者

Use this skill when the user requests current information, documentation, coding examples, or web research beyond the knowledge cutoff. Queries Google's AI Search mode to retrieve comprehensive AI-generated overviews with source citations from 100+ websites. Returns markdown with inline footnoted references [1][2][3]. You will receive a detailed Markdown file with references and information summarized directly from Google's AI search. Ideal for you to get new information and clues for further research.

2026-03-19
copilot-image-find-and-download
ソフトウェア開発者

Find and download images related to a user prompt through Copilot/Bing image discovery without requiring login. Use when the user asks for a prompt-to-image result, wants a quick downloadable image file, or specifically asks for Copilot image find-and-download behavior with output saved to ~/Downloads/temp.

2026-03-13
unsplash-image-search-download
ソフトウェア開発者

Search Unsplash for images related to a natural-language user prompt and download a selected result to disk using the official API and tracked download endpoint. Use when Codex needs stock photos from Unsplash for articles, slide decks, social posts, mockups, inspiration boards, or any prompt-to-image download workflow.

2026-03-13
protected-site-scraper
ソフトウェア開発者

Scrape websites protected by Cloudflare, anti-bot systems, and other WAF protections using undetected-chromedriver. Includes support for complex extraction, CAPTCHA handling, and data export. Works with sahibinden.com and similar protected real estate/classified sites.

2026-03-09
write-review-execute
ソフトウェア開発者

Generate and compare exactly three code scaffolds through a write-review-execute loop, then recommend the best solution using weighted multi-criteria scoring. Use when Codex must try alternative implementations, review likely failure points, optionally run proof-of-concept execution, and propose a defended best approach.

2026-03-09
coding-momentum-coach
運動トレーナー・グループフィットネスインストラクター

Inspect same-day Codex session activity, score coding momentum across the day, detect idle drop-off, and coach the next small mission to improve consistency. Use when a user asks for a coding day status, momentum check, re-entry prompt, daily streak review, or coaching based on intra-day sessions.

2026-03-09
humanize-text
作家・著者

Rewrite AI-sounding text so it reads naturally, specifically, and credibly without changing the underlying meaning. Use when users ask to humanize text, make writing less robotic, remove AI patterns, soften corporate phrasing, preserve facts while improving flow, or revise copy that was flagged by an AI detector.

2026-03-09
scrapling-websearch-fetch-data
ソフトウェア開発者

Fetch websites and extract structured data with Scrapling for research and agency workflows. Use when Codex needs to scrape pages, collect data from search result pages, crawl pagination, parse static or dynamic content, or build repeatable web data collection using Scrapling MCP tools, fetchers, and sessions.

2026-03-09
docx-create-edit
秘書・行政アシスタント(法律・医療・役員除く)

Create new DOCX files from prompts and update existing application-form, formal-request, or similar professional DOCX files with AI-drafted sections while preserving structure, tone, and file safety. Use when Codex must read a `.docx`, draft content such as qualifications, supporting statements, examples, motivation paragraphs, or formal responses, insert or replace targeted sections, or generate a fresh document from scratch.

2026-03-09
goal-constraint-inference
プロジェクト管理専門家

Infer hard constraints, negative constraints, optimization constraints, dependency/conflict relations, and priority weights from high-level goals, then produce alternative solution portfolios and a best balanced recommendation. Use when a user asks to turn a target into feasible constraints, clarify what must not happen, resolve conflicting requirements, compare tradeoffs, define what goal should be optimized, or apply forward and inverse optimization reasoning.

2026-03-09
stitch-remotion-walkthrough
ソフトウェア開発者

Build Stitch-based walkthrough videos and presentation-ready screen sequences with Remotion, including asset staging, screen manifests, modular compositions, and render validation. Use when Codex must turn Stitch project screens or exported screenshots into a demo video, promo walkthrough, app tour, or other screen-driven visual artifact.

2026-03-09
nano-banana
ソフトウェア開発者

Generate and iteratively refine images with the Gemini CLI Nano Banana extension. Use when Codex or OpenCode needs to create, edit, restyle, restore, or vary visual assets such as hero images, thumbnails, icons, diagrams, mockups, illustrations, patterns, or photo edits.

2026-03-09
code-documentation
ソフトウェア開発者

Create, rewrite, and review code documentation for repositories, modules, APIs, functions, and engineering workflows. Use when Codex must add or improve README sections, architecture notes, inline code docs, onboarding guides, usage docs, or review existing technical documentation for accuracy, gaps, and maintainability.

2026-03-07
crawl4ai-coder-helper
ソフトウェア開発者

Build, extend, and debug Crawl4AI-based scrapers, crawlers, and small services. Use when Codex needs to choose how to use Crawl4AI for a target website, inspect the site, pick an extraction strategy, and write the actual project files such as scraper modules, service endpoints, configs, schemas, dependency files, and sample outputs instead of only giving guidance.

2026-03-07
playwright
ソフトウェア品質保証アナリスト・テスター

Build, debug, and maintain Playwright end-to-end browser tests for web workflows. Use when Codex needs to create or update Playwright tests, fix flaky E2E behavior, configure Playwright projects, improve selectors and fixtures, or verify critical user journeys across browsers.

2026-03-05
analytics-tracking
市場調査アナリスト・マーケティングスペシャリスト

Design or audit product and marketing analytics instrumentation for reliable decision-making. Use when users ask for event taxonomy, GA4 or GTM setup, funnel tracking, attribution instrumentation, analytics QA, or reporting models tied to growth goals.

2026-03-05
content-strategy
市場調査アナリスト・マーケティングスペシャリスト

Plan a practical content strategy that connects business goals to topics, formats, and distribution. Use when users ask what to publish, how to build content pillars, how to prioritize ideas, or how to create a roadmap for organic and demand generation growth.

2026-03-05
copywriting
作家・著者

Write or improve conversion-focused marketing copy for websites and campaigns. Use when users request homepage or landing page copy, value proposition work, CTA optimization, offer messaging, or rewriting weak copy to improve clarity and conversion.

2026-03-05
marketing-ideas
市場調査アナリスト・マーケティングスペシャリスト

Generate and prioritize practical marketing ideas for software and SaaS growth. Use when users ask for growth ideas, channels to test, campaign brainstorming, or what to do next when marketing momentum is stalled.

2026-03-05
marketing-psychology
市場調査アナリスト・マーケティングスペシャリスト

Apply behavioral science and decision psychology to messaging, offers, and growth experiments. Use when users ask why customers do not convert, how to reduce decision friction, how to improve persuasion ethically, or how to use mental models in marketing strategy.

2026-03-05
postbridge-social-growth
市場調査アナリスト・マーケティングスペシャリスト

Coach organic short-form social growth on TikTok and Instagram using a structured warmup, content-market-fit, conversion, and scaling process. Use when users ask how to grow views organically, turn social traffic into installs or customers, fix low-view accounts, or scale repeatable short-form content systems.

2026-03-05
seo-audit
市場調査アナリスト・マーケティングスペシャリスト

Diagnose SEO performance problems and produce an actionable technical and on-page remediation plan. Use when users ask for an SEO audit, ranking drop analysis, indexing or crawl debugging, Core Web Vitals review, or broad organic growth diagnostics for a site or set of pages.

2026-03-05
web-deepsearch
ソフトウェア開発者

Perform deep web research for complex questions that require decomposition, multi-query coverage, evidence ranking, contradiction handling, and practical recommendations with citations. Use when the user asks to compare options, evaluate tradeoffs, design architecture or integration strategy, select tools or vendors, or needs high-confidence research beyond quick lookup.

2026-03-05
prompt-improvement
ソフトウェア開発者

Transform weak prompts into high-performance prompts for OpenAI and Codex workflows by asking targeted clarifying questions, preserving intent, and producing explicit structured prompts with testable constraints, verification expectations, and reliable output schemas.

2026-03-05
phase-planning
プロジェクト管理専門家

Create and maintain project planning documents in the project-root `docs/` directory by generating and synchronizing a master plan with phase plans in either split files (`docs/phases/phase-XX-slug.md`) or a consolidated roadmap file (`docs/PHASES.md` or existing equivalent). Use when Codex must break work into phases, update phase status/scope/dependencies, or keep master and phase artifacts aligned.

2026-03-05
document-creator
テクニカルライター

Create structured documents from selected templates and render PDF outputs with deterministic layout rules, section ordering, and reference formatting. Use when Codex must generate formal papers or reports from user content, ask for template selection first, enforce a chosen template, and produce a final document file.

2026-03-04
create-guidance
マネジメントアナリスト

Clarify vague or high-level requests, propose options, and give direct step-by-step guidance. Use when a user asks for help understanding what to do, how to do it, or needs intent clarification to turn a goal into actionable steps.

2026-03-04
tdd
ソフトウェア品質保証アナリスト・テスター

Build features and fix bugs using a strict Test-Driven Development workflow (Red-Green-Refactor), with regression-first bugfixing and explicit clarification when expected behavior is ambiguous. Use when Codex must implement or repair behavior with high confidence, test evidence, and minimal-risk changes.

2026-03-04
solution-alternatives
ソフトウェア開発者

Generate and compare multiple solution alternatives for hard issues using explicit constraint modeling, dependency and conflict analysis, and optimization-function reasoning. Use when Codex must evaluate tradeoffs, handle ambiguity with clarifying questions, refine problematic constraints, and recommend a primary option with fallback conditions.

2026-03-04