resumegenerator

Markdown-to-PDF resume generator. Write your resume in markdown, get a styled, ATS-friendly PDF.

Install

Requires Bun.

bun i -g @mpelka/resumegenerator

This installs the CLI tool and Chromium browser automatically. Fonts are downloaded from Google Fonts on first run and cached locally.

Usage

resumegenerator --filename resume.md

| Flag | Description | |------|-------------| | --filename | Path to the markdown resume (required) | | --template | Template name (default: modern) | | --initials | Override auto-derived monogram initials | | --output-filename | Override the output PDF filename | | --spacing | Spacing multiplier (e.g. 0.8 = 80% of default gaps) |

The PDF is written next to the source .md file. Initials for the monogram are auto-derived from the h1 name.

Markdown format

# Full Name

**Job Title**

Location • [email protected] • linkedin.com/in/handle

Summary paragraph with optional **bold** keywords.

## Work Experience

### Company Name
**Role Title** | Start Date - End Date

* Achievement with **keyword** highlights

*Technologies used: Tech1, Tech2, Tech3.*

## Education

### University Name
**Degree** | Start Year - End Year

## Skills

- **Category:** Item1 • Item2 • Item3

YAML frontmatter is supported and stripped before rendering.

Templates

Each template defines its own fonts, colors, and layout features. Pass --template <name> to switch.

| Template | Description | |----------|-------------| | modern (default) | Clean sans-serif (IBM Plex Sans) with monogram badge and section divider lines | | technical | Monospace (IBM Plex Mono) with outlined circle monogram and minimal section labels |

Templates live in templates/<name>/ with two files:

• style.css — layout, margins (@page), and visual styling
• template.js — config (fonts, colors, feature flags)

Customizing templates

To create a new template, add a directory under templates/ with:

template.js — exports a config object:

export default {
  fonts: { primary: "IBM Plex Sans", secondary: "IBM Plex Mono" },
  colors: {
    body: "#323336",
    subtitle: "#707678",
    sectionLabel: "#a6aaad",
    accent: "#42f398",
    border: "#e0e0e0",
  },
  features: {
    monogram: true,        // render initials badge
    sectionDividers: true, // "EXPERIENCE ————" style dividers
  },
};

style.css — template-specific styles, including page margins via @page. Uses CSS variables (--font-primary, --font-secondary, --color-*) injected from the config at build time.

Font families must be present in the GOOGLE_FONTS_URL map in src/utils.ts.

How it works

The CLI parses your markdown, renders it to semantic HTML with configurable templates, downloads and caches Google Fonts as TTF files, then uses Playwright's Chromium to generate a tagged PDF. The result is an ATS-optimized document with proper text selection, heading structure, and PDF link annotations.

ATS optimization

The generated PDFs are optimized for Applicant Tracking Systems and AI-based resume screening:

• Tagged PDF — embeds a structural tag tree (headings, paragraphs, lists) for correct semantic parsing
• PDF metadata — title is set to Name - Resume (not the temp filename)
• Single-column layout — no tables, columns, or complex layouts that break ATS parsers
• Real text — all content is selectable text, not images
• Semantic HTML — clean heading hierarchy (h1/h2/h3) with standard section names
• Hyperlinks — email, LinkedIn, and GitHub are embedded as proper PDF link annotations

Font strategy

The CLI downloads full TTF files from Google Fonts rather than using CDN <link> tags. Google Fonts CDN serves WOFF2 files split into unicode-range subsets, and when Playwright embeds these into a PDF the character-to-glyph mapping fragments — text looks correct but copy-paste produces garbled output, breaking ATS parsers. Fetching with a basic Linux User-Agent returns un-subsetted TTF URLs that produce clean character maps. A temp HTML file is written so the page loads with a file:// origin, which is required for Chromium to access the locally cached font files.