# parser > Instructions for the Agent to convert Markdown to HTML fragment manually. - Author: Paul - Repository: paulkwongemini/cila-translate - Version: 20260129103033 - Stars: 0 - Forks: 0 - Last Updated: 2026-02-06 - Source: https://github.com/paulkwongemini/cila-translate - Web: https://mule.run/skillshub/@@paulkwongemini/cila-translate~parser:20260129103033 --- --- name: parser description: Instructions for the Agent to convert Markdown to HTML fragment manually. --- # Parser Skill (Agent Instruction) This skill provides the **Standard Operating Procedure (SOP)** for the Agent to convert the translated Markdown file into the final HTML fragment without using a script. ## Procedure 1. **Read Input**: - Read `workspace/translated.md` (or specified file). 2. **Filter & Prepare**: - **Ignore Top Title**: If the first non-empty line starts with `#` and contains "주간 기도문", ignore it (do not include in output). - **Filter English**: If the document contains interleaved English and Korean, **exclude all English content**. Keep only the Korean sections. 3. **Apply HTML Conversion**: - **Headers**: Convert lines starting with `#` to `

`. Remove bold `` markers if present. - Lists**: Convert lines starting with `* ` (asterisk + space) into `
` with `
`. Combine consecutive list items into a single `
`. - Paragraphs: Wrap other non-empty lines in `
`. - Links: Convert `[text](url)` to `text`. 4. Final Assembly: - Do not create a full HTML skeleton (``, ``). - Prepend the CSS Block (see below). - Wrap the entire converted content in a `
` container. - Ensure there is no newline between the CSS block and the `
`. 5. Output: - Save the result to `outputs/final.html`. ## Reference: CSS Block Use this exact code block at the top of the file: ```html ```