GitHub has this concept of profile README. It’s basically a Markdown document that shows up on your public GitHub profile. It works by having a README.md file in a repository named after your GitHub username. For instance, my GitHub username is KittyGiraudel, so the README.md file of my KittyGiraudel repository would get displayed (see it in action).
It would be nice if this file could list my latest articles, so that it links back to my website for people to discover my content. But also I shouldn’t have to manually update that file because it’s cumbersome and we live in the age of automation.
Someone on LinkedIn shared that they asked Claude to set up a CRON job, and that got me thinking. GitHub Actions support CRON workflows, so that’s probably all we need?
How it works
Here is the gist of it: the README.md file contains a fenced section that gets wiped and replaced with recent articles. A Node.js script fetches my RSS feed, grabs the last entries, and generates a Markdown table to list them. It then injects it in the README.md file. Finally, a GitHub Action runs that script and push a commit to update the file on the remote repository.
Templating the README
The nice thing about Markdown is that we can use HTML comments. This way, they are invisible when rendering the file. We put a placeholder sentence in between them so that if the script fails for any reason, the content remains meaningful.
## Latest writing
<!-- BLOG-POST-LIST:START -->
Find my latest blog posts on [kittygiraudel.com](https://kittygiraudel.com/blog).
<!-- BLOG-POST-LIST:END -->
Creating the Node.js script
Our script needs to do 3 different things:
- Get the latest articles.
- Generate a Markdown table to list them.
- Inject that table in the
README.mdfile.
Retrieving latest articles
First, let’s fetch the latest articles. It’s made very easy by using my RSS feed and a small library to parse XML:
import { XMLParser } from 'fast-xml-parser'
async function fetchFeed() {
try {
// Fetch the RSS feed
const response = await fetch('https://kittygiraudel.com/rss/index.xml')
if (!response.ok) {
console.error(
</span><span class="token string">Failed to fetch feed: </span><span class="token interpolation"><span class="token interpolation-punctuation punctuation">${</span>response<span class="token punctuation">.</span>status<span class="token interpolation-punctuation punctuation">}</span></span><span class="token string"> </span><span class="token interpolation"><span class="token interpolation-punctuation punctuation">${</span>response<span class="token punctuation">.</span>statusText<span class="token interpolation-punctuation punctuation">}</span></span><span class="token template-punctuation string">,
)
return null
}
// Parse the resulting XML
const xml = await response.text()
const parser = new XMLParser({ ignoreAttributes: false })
const data = parser.parse(xml)
// Ensure we have some data to work with
const entries = data?.feed?.entry
if (!entries || (Array.isArray(entries) && entries.length === 0)) {
console.error('No entries found in feed.')
return null
}
// For each entry, resolve the necessary data (name, date, link)
const items = Array.isArray(entries) ? entries : [entries]
return items.slice(0, 5).map(entry => {
const title = entry.title ?? 'Untitled'
const url = entry.link?.['@_href']
const date = new Date(entry.published).toLocaleDateString('en-GB', {
month: 'short',
year: 'numeric',
})
return { title, url, date }
})
} catch (error) {
console.error('Error while fetching or parsing feed:', error)
return null
}
}
Generating Markdown table
The second thing we need is to generate the Markdown table for our data:
function buildMarkdown(posts) {
if (!posts || posts.length === 0) {
return 'Find my latest blog posts on [kittygiraudel.com](https://kittygiraudel.com/blog).'
}
const rows = posts
.map(post => {
const link = post.url || 'https://kittygiraudel.com'
const title = post.title.replace(/|/g, '\|')
return </span><span class="token string">| </span><span class="token interpolation"><span class="token interpolation-punctuation punctuation">${</span>post<span class="token punctuation">.</span>date<span class="token interpolation-punctuation punctuation">}</span></span><span class="token string"> | [**</span><span class="token interpolation"><span class="token interpolation-punctuation punctuation">${</span>title<span class="token interpolation-punctuation punctuation">}</span></span><span class="token string">**](</span><span class="token interpolation"><span class="token interpolation-punctuation punctuation">${</span>link<span class="token interpolation-punctuation punctuation">}</span></span><span class="token string">) |</span><span class="token template-punctuation string">
})
.join('\n')
return [
'| Date | Post |',
'| ---- | ---- |',
...rows,
].join('\n')
}
Updating README
Finally, we want to update the README.md file.
const MARKER_START = '<!-- BLOG-POST-LIST:START -->'
const MARKER_END = '<!-- BLOG-POST-LIST:END -->'
function updateReadmeBlock(content, block) {
const startIndex = content.indexOf(MARKER_START)
const endIndex = content.indexOf(MARKER_END)
if (startIndex === -1 || endIndex === -1 || endIndex <= startIndex) {
console.error('Could not find blog post markers in README.md')
return content
}
const before = content.slice(0, startIndex + MARKER_START.length)
const after = content.slice(endIndex)
return </span><span class="token interpolation"><span class="token interpolation-punctuation punctuation">${</span>before<span class="token interpolation-punctuation punctuation">}</span></span><span class="token string">\n</span><span class="token interpolation"><span class="token interpolation-punctuation punctuation">${</span>block<span class="token interpolation-punctuation punctuation">}</span></span><span class="token string">\n</span><span class="token interpolation"><span class="token interpolation-punctuation punctuation">${</span>after<span class="token interpolation-punctuation punctuation">}</span></span><span class="token template-punctuation string">
}
Running our script
We want to run our script on a regular basis (such as every day) to keep our profile README updated. We can create a GitHub workflow executing a CRON job.
name: Update profile README
on:
schedule:
- cron: '0 6 * * *' # Every day at 6am
permissions:
contents: write # Needed to let GitHub issue commits
jobs:
update-readme:
runs-on: ubuntu-latest
steps:
- name: Check out repository
uses: actions/checkout@v4
- name: Set up Node.js
uses: actions/setup-node@v4
with:
node-version: '24'
cache: 'npm'
- name: Install dependencies
run: npm install
- name: Update README with latest posts
run: npm run update-readme
- name: Commit and push changes
run: |
if git diff --quiet; then
echo "No changes to commit"
exit 0
fi
git config user.name "github-actions[bot]"
git config user.email "41898282+github-actions[bot]@users.noreply.github.com"
git add README.md
git commit -m "chore: update README with latest blog posts"
git push
We also want to run our workflow when pushing onto the main branch so that we can safely update our README.md template and have the script run.
on:
schedule:
- cron: '0 6 * * *'
push:
branches: [main]Caveats
One downside of issuing an automated commit is that our local branch ends up behind its remote counterpart, since there are some missing commits. We can of course pull to bring these commits in, but they end up containing a non-templated version of the README.md file, which is not really what we want.
I think a decent way is to just force push locally, to just ignore these generated commits. It’s questionable, but I have found that this works fine for me:
git add README.md
git commit -m "Update README.md template"
git push --forceWrapping up
Short and sweet, a fun 15-minute experiment. I hope it will inspire you to do something similar with your own GitHub profile! You can browse the code on GitHub.