Site-specific browser playbooks (domain-skills, interaction-skills) for Naver, GitHub, etc. Automate the browser with the browser-use CLI (state/indexed elements, eval, screenshots)โsame stack as skills/browser-use. Upstream browser-harness Python repo is optional for advanced CDP.
Amazon Bedrock ๊ธฐ๋ฐ PowerPoint(.pptx) ๋ฒ์ญ. ์์ยท๋ ์ด์์ยท์ฐจํธ ๋ฉํ(์ ๋ชฉ/์ถ ๋ฑ)๋ฅผ ๋ณด์กดํ๋ฉฐ ํ ์คํธ๋ฅผ ํ๊ตญ์ด(ko) ๋ฑ ๋์ ์ธ์ด๋ก ๋ณํ. CLI(`python -m ppt_translator.cli`)ยทSQLite ์บ์ยท์ฉ์ด์งยท ์๋ฌธ ์ธ์ด ์๋ ๊ฐ์ง. PPT/์ฌ๋ผ์ด๋ ๋ฒ์ญ, pptx ํ๊ตญ์ด, Bedrock ํ๋ ์ ํ ์ด์ ๋ฒ์ญ, batch translate, dry-run ๋น์ฉ ์ถ์ .
Create professional AWS-themed PowerPoint presentations with two theme options: dark (reInvent 2023/2025 with gradient backgrounds) and light (L100/field enablement with white backgrounds and soft gradient accents). Uses AWS brand colors, rich visual elements (SVG diagrams, architecture diagrams, icons), and cross-skill integration (svg-diagram). Supports creating from scratch, editing specific slides conversationally, and embedding SVG visualizations for maximum design impact. Trigger: "myslide", "make slides", "AWS presentation", "create pptx", "slide deck", "AWS slides", "ํ๋ ์ ํ ์ด์ ", "์ฌ๋ผ์ด๋ ๋ง๋ค์ด", "๋ฐํ์๋ฃ", "L100", "L200", "training deck", "light theme", "๋ฐ์ ํ ๋ง", "๊ต์ก ์๋ฃ"
ํด๋ ๋ด ๋ชจ๋ ์ด๋ฏธ์ง๋ฅผ AWS Bedrock(LLM) ๋ฉํฐ๋ชจ๋ฌ๋ก ์์ฐจ ๋ณํํด ํ๋์ ํตํฉ Markdown ํ์ผ(.md)๋ก ๋ง๋ ๋ค. `mcp_server_text_extraction.py`์ ๋์ผํ ์ ์ฒ๋ฆฌ(ํฌ๊ธฐยทํฝ์ ์ ํ)ยท์ถ์ถ ํ์ดํ๋ผ์ธ์ ์ฌ์ฉํ๋ค. ์ถ๋ ฅ์ ๋ช ์์ ์ผ๋ก Markdown์ด๋ฉฐ, ๋ํยท๊ทธ๋ฆผ ๋ฑ์ด ์์ผ๋ฉด ์๋ฏธ๋ฅผ ์์ธ ์์ ํ๋๋ก ํ๋กฌํํธํ๋ค. ์๋จ headerยทํ๋จ footer ์ ์ธ ๋ฐ ๋ฌธ์ฅ ๋จ์ ๊ตฌ๋ถ์ ์ ์ฉํ๋ค. ํธ๋ฆฌ๊ฑฐ: img2text, ์ด๋ฏธ์ง ํด๋์์ Markdown ์ถ์ถ, ๋ฉํฐ ํ์ด์ง ์ด๋ฏธ์ง LLM ๋ณํ, `artifacts/...` ๊ฐ์ ์ค์บ ํด๋ ์ ์ฒด LLM ์ถ์ถ, WB_Troubleshooting ๊ฐ์ ๋งค๋ด์ผ ํด๋ ์ผ๊ด ๋ณํ.
Extract text from an image file (PNG, JPG, etc.) using Tesseract OCR and save the result as a .txt file. Automatically crops the top header and bottom footer regions before OCR so that page titles, section headings in the margin, and page numbers are excluded from the output. Use this skill whenever the user wants to convert an image to text, extract body text from a scanned page, or perform OCR on a document image while skipping headers and footers (e.g. "page_033.png์ ํ ์คํธ๋ก ๋ณํํด์ค", "์ด ์ด๋ฏธ์ง์์ ๋ณธ๋ฌธ๋ง ์ถ์ถํด์ค", "img2text๋ก ๋ณํํด์ค").
Convert a PDF file into per-page PNG images and save them into a dedicated subfolder inside the artifacts directory. Use this skill whenever the user provides a PDF path (e.g. ~/Downloads/6660-SPS.pdf) and asks to extract, convert, or save its pages as images. The output folder is named after the PDF file stem (e.g. artifacts/6660-SPS/).
Generate a account status report by taking an account name, analyze spend trends and AWS account mappings, create an HTML report, render chart image for email compatibility, ask recipient email after report completion, and send immediately without reconfirmation. All analysis and email narrative must be in Korean.
AWS Bedrock์์ Anthropic Claude ๋ชจ๋ธ์ ๋ฆฌ์ ๋ณ ์ง์ ์ฌ๋ถ๋ฅผ ํ์ธํ๋ ์คํฌ. ์ฌ์ฉ์๊ฐ "sonnet 4.7", "opus 4.7", "haiku 4.5" ๋ฑ์ ๋ชจ๋ธ๋ช ์ ์ ๋ ฅํ๋ฉด AWS ์ฃผ์ 10๊ฐ ๋ฆฌ์ (๋ฏธ๊ตญ, ์ ๋ฝ, ์์์ ํํ์)์์ ํด๋น ๋ชจ๋ธ์ด ์ง์๋๋์ง ์กฐํํ๊ณ ์ง์ ๋ฆฌ์ ๋ชฉ๋ก๊ณผ ์ค์ ๋ชจ๋ธ ID๋ฅผ ์๋ ค์ค๋ค. ๋ค์๊ณผ ๊ฐ์ ์์ฒญ์ ์ฌ์ฉํ๋ค: (1) "sonnet 4.7 ์ด๋ ๋ฆฌ์ ์์ ์ธ ์ ์์ด?", (2) "opus 4.7 ์ง์ ๋ฆฌ์ ์๋ ค์ค", (3) "AWS Bedrock์์ claude haiku 4.5 ์ธ ์ ์๋ ๋ฆฌ์ ์?", (4) ํน์ Claude ๋ชจ๋ธ์ ๊ฐ์ฉ ๋ฆฌ์ ํ์ธ ์์ฒญ