A starter template for rendering your own resume with
ferrocv — the JSON Resume →
PDF/HTML/plain-text renderer powered by Typst.
A minimal, forkable repo that takes a single resume.json file and
renders it to PDF, HTML, and plain text on every push via GitHub
Actions, then publishes the results as a rolling latest GitHub
Release.
- Input:
resume.jsonat the repo root (JSON Resume v1.0.0). - Output:
resume.pdf,resume.html, andresume.txtattached to alatestrelease that is overwritten on every push tomain. Keep the repo private and the release (and its assets) stay private too. Every run (including PRs) also uploads the renders as a 14-day workflow artifact (resume-<sha>) for preview without publishing. - Checks: Every push and PR spell-checks with
typos, schema-validates viaferrocv validate, and link-checks URLs withlycheebefore rendering. - Runtime: a pinned
ferrocvbinary downloaded from the upstream GitHub Release. No Rust toolchain, no Node, no TeX.
The sample resume.json describes a fully fictional person. Replace it
with your own.
- Click "Use this template" → "Create a new repository" at the top of this repo on GitHub. Keep it private if you don't want your resume publicly browsable.
- Edit
resume.jsonwith your own content. The schema reference lives at https://jsonresume.org/schema/. - Commit and push to
main. Thebuildworkflow validates your resume, renders it, and updates thelatestrelease. PRs tomainrun the same checks and upload a preview artifact without publishing — handy for reviewing changes before they go live. - Grab your rendered PDF from the repo's Releases → latest page. The download URL is stable; on a private repo you'll need to be signed in to GitHub to fetch it.
Pages is off by default. Turning it on means your resume is served as a website — and on Free/Pro/Team plans, Pages from a private repo is still published to a public URL. Only GitHub Enterprise Cloud supports private-access Pages.
If you still want it, run the workflow manually:
Actions → build → Run workflow, set
Also deploy to GitHub Pages to true. Site will appear at
https://<your-username>.github.io/<your-repo>/ after you set
Settings → Pages → Source to GitHub Actions.
Install ferrocv (see the
releases or
cargo install ferrocv), then:
ferrocv validate resume.json
ferrocv render resume.json --theme classic --output dist/resume.pdf
ferrocv render resume.json --format html --output dist/resume.html
ferrocv render resume.json --format text --output dist/resume.txtRun ferrocv themes list to see the themes shipped with your
installed version.
The prompts/ directory ships three starter prompts for drafting,
reviewing, and tailoring your resume.json with any AI assistant:
builder.md— schema-aware drafting: turn a brain-dump, LinkedIn export, or rough bullets into valid JSON Resume fragments.reviewer.md— content audit: flags weak verbs, missing metrics, passive voice, buzzword stacking, and schema smells.tailor.md— job-posting matcher: surfaces which existing content to emphasize, de-emphasize, and honestly-missing gaps for a specific role. Never invents experience.
They're written as plain markdown so they travel across assistants
(Claude, ChatGPT, Gemini, Copilot, Cursor, Aider, etc.) and between
chat and coding UIs. Drop the file contents into a system prompt, a
project instructions field, or a tool-specific wrapper
(.cursor/rules/, .github/copilot-instructions.md, a Claude Code
slash command, etc.) — whichever your workflow prefers.
Treat them as a starting point, not a spec. Fork and evolve to fit your voice and the tools you actually use.
The seed workflow renders a PDF with classic (ferrocv's built-in
default theme), plus HTML and plain text via the default text-minimal
theme. Other PDF themes (typst-jsonresume-cv, modern-cv,
fantastic-cv) ship with ferrocv — swap the --theme value or add
extra render steps. Output filenames should
follow the resume.<format> convention (resume.pdf, resume.html,
resume.txt) so they all land cleanly on the latest release.
ferrocv is installed via the cacack/ferrocv/.github/actions/setup-ferrocv
composite action, pinned to a commit SHA. To update, pick a newer tag
from the ferrocv releases page
and edit .github/workflows/build.yml in four places total (two jobs ×
two lines each):
- the action's pinned SHA (the
@<sha>ref, with the# vX.Y.Zcomment updated to match the new tag), and - the matching
version:input.
Resolve the SHA for a tag with
gh api repos/cacack/ferrocv/git/refs/tags/<tag>.
The spell- and link-checks will occasionally flag things that aren't actually wrong (proper names, rate-limited hosts, intentionally internal URLs). Both tools read optional config from the repo root:
- Typos: add a
.typos.toml— see the typos configuration docs for the[default.extend-words]escape hatch. - Lychee: add a
lychee.toml— see the lychee configuration docs forexcludepatterns andacceptstatus codes.
This template is dedicated to the public domain under
CC0 1.0 Universal. Your own resume.json content is
yours — this dedication applies only to the template scaffolding.