SFA Jobs — jobs.streetsforall.org

Public-sector job board for the Streets For All community. Every morning it pulls open positions from LA-area agencies, keeps the ones in planning/engineering, outreach/policy, GIS/data and project management, publishes them at jobs.streetsforall.org, and posts anything new to #jobs in Slack.

Where everything runs

Job	Runs on	Triggered by	Code
Fetch listings, post new ones to Slack #jobs, commit site/jobs.json	GitHub Actions jobs	schedule, daily ~6:40am PT	src/run.js
Job board page	Squarespace page on streetsforall.org (Code Block) that loads embed.js	—	site/embed.js
jobs.json + embed.js host (jobs.streetsforall.org)	GitHub Pages (deployed by the same workflow)	each run, and pushes that change site/	site/

Secrets and variables (names only):

• SLACK_BOT_TOKEN (secret) — the existing Streets For All Slack app, same one agenda-watch uses. The bot must be a member of #jobs.
• SLACK_JOBS_CHANNEL_ID (variable) — C08CFG5RFQ8 (#jobs).

The Squarespace page

The board is a normal Squarespace page (e.g. streetsforall.org/jobs) with one Code Block:

<div id="sfa-jobs"></div>
<script src="https://jobs.streetsforall.org/embed.js" defer></script>

The widget inherits the site's fonts and uses the S4A palette; the page title and intro are ordinary Squarespace text blocks. jobs.streetsforall.org itself redirects to that page (REDIRECT in site/index.html); add ?preview to see the standalone version.

Agencies and sources

See config/agencies.js.

Agency	Source
City of LA (incl. LADOT, LADWP, StreetsLA), LA County, LA Metro, Metrolink, SCAG	NEOGOV job feed
Long Beach, Pasadena, Santa Monica (incl. Big Blue Bus), Glendale, Burbank, Culver City (incl. CityBus), West Hollywood, Beverly Hills, Torrance (incl. Torrance Transit)	NEOGOV job feed
Caltrans District 7	CalCareers search (Department of Transportation, Los Angeles County)
City of Inglewood	CivicPlus jobs RSS
Long Beach Transit	Oracle Cloud HCM careers site

NEOGOV is NEOGOV's official RSS job feed (governmentjobs.com/SearchEngine/JobsFeed?agency=<slug>), the syndication feed NEOGOV documents for agencies to publish openings to outside sites. The board reads only that feed; it does not crawl governmentjobs.com pages.

CalCareers has no feed or API, so src/fetchers/calcareers.js submits the public search form the way a browser does. It is the most fragile source; if it breaks, only Caltrans stops refreshing.

Not covered: Foothill Transit takes emailed PDF applications and its operations staff are hired by contractors, so there is nothing to pull.

A source that fails on a given day keeps its previous listings on the site and is shown in red in the site footer; the Actions run summary lists every source's status.

What gets listed

Decided from the job title in config/categories.js:

• matches a category pattern (engineering & planning, outreach & policy, GIS & data, project management), or
• is a generic analyst/manager title in a transportation, public works or planning department, or
• contains a transportation term (active transportation, bike, pedestrian, transit, mobility…) with a professional title.

Excluded: police/fire/health/library and similar job families, IT engineering, and frontline roles such as bus operators, mechanics and traffic officers. Promotional (current-employees-only) recruitments and closed postings are dropped.

To adjust, edit the patterns, add a case to test/classify.test.js, and run npm test.

Running it

npm ci
npm test                          # classifier, feed parser, Slack message
node src/run.js --dry-run         # fetch + print new jobs, write nothing
node src/run.js --only=lacity --dry-run
node src/run.js --baseline        # mark everything open as seen, no Slack post

The same baseline and dry_run options are on the workflow's "Run workflow" button.

Launch checklist

1. Create the repo streetsforall/jobs and push this code.
2. Settings → Pages → Source: GitHub Actions. Custom domain: jobs.streetsforall.org, enforce HTTPS.
3. DNS: add CNAME jobs → streetsforall.github.io.
4. Add secret SLACK_BOT_TOKEN and variable SLACK_JOBS_CHANNEL_ID=C08CFG5RFQ8; /invite the bot to #jobs.
5. Run the workflow with dry_run and check each source in the run summary.
6. Run it once with baseline (publishes the site, no Slack post).
7. From then on the daily schedule posts only genuinely new jobs.