hud.ac.uk logo
hud.ac.uk

find-program-details

Installation

Adds this website's skill for your agents

 

Summary

Retrieve details for University of Huddersfield academic programs — list every course or fetch one course's level, duration, start dates, delivery mode, entry requirements, fees, and modules — via the coursefinder's SearchStax/Solr JSON API, with a prerendered-page and browser fallback.

FIG. 01
FIG. 02
FIG. 03
FIG. 04
SKILL.md
163 lines

Find University of Huddersfield Program Details

Purpose

Retrieve details for academic programs (courses) offered by the University of Huddersfield — either a full listing of every course or the details of a specific program. Returns per-course fields such as title, study level, academic year, duration, start dates, delivery mode/flags, the course URL, and (for a given course) the full overview, entry requirements, modules, careers, and tuition fee. This is a read-only skill. The fastest and most complete path is the site's own JSON search API (SearchStax/Solr emselect), which returns all 787 course records; individual course pages are prerendered HTML that can be fetched for the remaining structured fields (fee, UCAS tariff, module lists).

When to Use

  • "List every undergraduate/postgraduate course at Huddersfield" (optionally filter by year, level, or delivery mode).
  • "Get the details for a named program" (e.g. Computer Science BSc(Hons), 2026-27).
  • Comparing courses by duration, start date, study level, or delivery flag (Distance Learning, Apprenticeship, Foundation Year, etc.).
  • Building a dataset/export of the full Huddersfield course catalogue.
  • Looking up entry requirements, tuition fees, or module lists for a specific course.

Workflow

The University's coursefinder (https://courses.hud.ac.uk/) is a Vue SPA backed by a hosted SearchStax/Solr index. Do not scrape the SPA to enumerate courses — query the same JSON endpoint the app uses.

Recommended method — SearchStax emselect JSON API

  1. Endpoint & auth. The coursefinder ships its (client-side, non-secret) read token in the page source. Send it as a header — the token is not accepted as a query param.

    • URL: https://searchcloud-1-eu-west-2.searchstax.com/29847/huddersfielddevelopment-4954/emselect
    • Header: Authorization: Token 0887c7290da5f22a63dfbac3af780f99103df2fc
    • Method: POST (or GET) with Content-Type: application/x-www-form-urlencoded
    • If the token has rotated, re-scrape it: GET https://courses.hud.ac.uk/ and read config.searchAuth from the inline <script> (also exposes searchURL, suggesterURL, searchProfile: "Coursefinder").

  2. Restrict to courses. The index holds ~9,900 mixed docs (web pages, PDFs, news). Filter to course records only:

    • fq=sectionType_s:course → 787 course documents.

  3. Query / paginate. Standard Solr params apply:

    • List everything: q=*:*&fq=sectionType_s:course&rows=100&start=0&wt=json (page with start, rows).
    • Search by keyword: q=computer science (dismax over title/content).
    • Filter: add fq=study_level_s:Undergraduate, fq=year_s:2026-27, fq=flags_list_ss:"Distance Learning".
    • Trim the payload with fl=searchTitle_t,study_level_s,year_s,start_dates_s,duration_t,flags_list_ss,url,content.
    • Facet for catalogue structure: facet=true&facet.field=study_level_s&facet.field=year_s&facet.field=flags_list_ss.

  4. Read the results. Each doc in response.docs gives the summary fields plus a content field containing the full page text (overview, entry requirements, modules, careers). For most "program details" needs this single call is sufficient.

  5. Enrich a specific course (optional). For clean structured fields not broken out in Solr (tuition fee, UCAS tariff points, per-year module lists), the course page itself is prerendered HTML — fetch it directly (no JS execution needed):

    • GET {doc.url} e.g. https://courses.hud.ac.uk/2026-27/undergraduate/computer-science-bsc-hons/ returns fully-populated HTML with the fee (£9,790), UCAS tariff, A-Level/BTEC/T-Level entry criteria, and module lists.

Browser fallback (when the API token is unavailable or has changed and you can't re-scrape it)

  1. browse open https://courses.hud.ac.uk/ --remote (no stealth needed — see Gotchas).
  2. Dismiss the CiviC "Your choice regarding cookies" banner (click Accept / Agree) — it overlays the <h1>, so ignore h1 text for page identification.
  3. Deep-link a search instead of typing: https://courses.hud.ac.uk/?searchstax[query]=computer%20science&searchstax[model]=Coursefinder (URL-encode the brackets).
  4. Click the target course card to open its detail page (URL pattern /{year}/{level}/{slug}/).
  5. Extract with browse get text body — not browse snapshot. Wait ~2s (browse wait timeout 2000) after navigation for the Vue app to hydrate before reading.

Site-Specific Gotchas

  • Anti-bot: none. hud.ac.uk and courses.hud.ac.uk returned no bot protection across the whole run. A bare Browserbase session (no --verified, no --proxies) works; the successful runs used neither. The SearchStax endpoint is a third-party host with no bot protection either.
  • Auth is header-only. emselect returns 401 {"message":"Unauthorized"} with no token and when the token is passed as a ?token= query param. It only accepts Authorization: Token <searchAuth>. browse cloud fetch cannot send custom headers, so run the API call from an HTTP client that can (curl/python/node), or from inside the page context via browse eval (same-origin, CORS is open to courses.hud.ac.uk).
  • The Solr collection is named huddersfielddevelopment-4954 — this is genuinely what the production coursefinder queries; don't assume it's a staging index you shouldn't use. That said, a small number of not-yet-launched records (e.g. some Clearing-only courses) contain placeholder text like testtesttest / Content / a good job 3sgdsdg in their content field. Treat obviously templated content as incomplete rather than real.
  • Always filter fq=sectionType_s:course. Without it you get news, PDFs, and general web pages. Facet values on type_s are file types (web page/pdf/docx), not course levels.
  • Duplicate records per academic year. Most courses appear once per year (year_s = 2026-27 and 2027-28), so a course like Computer Science MSc shows up twice with different URLs. Dedupe on title or filter fq=year_s:<year> if you want one row per course. Totals: 787 records = 387 (2026-27) + 394 (2027-28) + a few Clearing/other.
  • study_level_s casing. Real values are Undergraduate (422) and Postgraduate (365). Lowercase undergraduate/postgraduate buckets exist in facets but are empty — match the capitalised forms.
  • Delivery modes live in flags_list_ss (multi-valued): On-Campus, Distance Learning, Short Course, Top-Up, Apprenticeship, Foundation Year, Teacher Training, Blended Learning, Research Degree.
  • Course detail pages are prerendered, so browse cloud fetch {course_url} returns the full content (fee/UCAS/entry text) without needing to render JS — but the coursefinder home/search results are client-rendered and must be reached via the API or a hydrated browser.
  • In the browser, prefer browse get text body over browse snapshot. During testing browse snapshot intermittently failed (returned a CLI "update available" notice instead of the tree); text extraction was reliable. Wait ~2s for Vue hydration first.
  • Cookie banner overrides <h1>. Until dismissed, document.querySelector('h1') returns "Your choice regarding cookies on this site" — use the URL or page title to confirm you're on the right course, not the h1.
  • There is also an AWS API Gateway (https://8dg66giahj.execute-api.eu-west-2.amazonaws.com/ProdStage/) referenced only as a hidden honeypot link in the DOM — do not call it; it is bait for scrapers, not a data source.

Expected Output

Listing shape (from the emselect API, one object per course record):

{
  "total_courses": 787,
  "query": { "q": "*:*", "fq": ["sectionType_s:course", "year_s:2026-27"] },
  "courses": [
    {
      "title": "Computer Science BSc(Hons)",
      "study_level": "Undergraduate",
      "year": "2026-27",
      "duration": "3 years full-time 4 years inc. placement year",
      "start_dates": "21 September 2026",
      "delivery_flags": ["On-Campus"],
      "url": "https://courses.hud.ac.uk/2026-27/undergraduate/computer-science-bsc-hons/"
    },
    {
      "title": "Computer Science MSc",
      "study_level": "Postgraduate",
      "year": "2026-27",
      "duration": "1 year full-time",
      "start_dates": "Multiple start dates",
      "delivery_flags": ["On-Campus"],
      "url": "https://courses.hud.ac.uk/2026-27/postgraduate/computer-science-msc/"
    }
  ],
  "facets": {
    "study_level_s": { "Undergraduate": 422, "Postgraduate": 365 },
    "year_s": { "2027-28": 394, "2026-27": 387 },
    "flags_list_ss": {
      "On-Campus": 698, "Distance Learning": 89, "Short Course": 81,
      "Top-Up": 53, "Apprenticeship": 48, "Foundation Year": 38,
      "Teacher Training": 24, "Blended Learning": 20, "Research Degree": 6
    }
  }
}

Single-program detail shape (API summary fields + fields enriched from the prerendered course page):

{
  "success": true,
  "title": "Computer Science BSc(Hons)",
  "study_level": "Undergraduate",
  "year": "2026-27",
  "duration": "3 years full-time 4 years inc. placement year",
  "start_dates": "21 September 2026",
  "delivery_flags": ["On-Campus"],
  "tuition_fee": "£9,790",
  "typical_entry_requirements": "BBB-BBC at A Level. DMM in BTEC Level 3 Extended Diploma. 120-112 UCAS tariff points from a combination of Level 3 qualifications. Merit at T Level. GCSE English at grade 4+ and Maths at grade 5+.",
  "url": "https://courses.hud.ac.uk/2026-27/undergraduate/computer-science-bsc-hons/",
  "error_reasoning": null
}

Not-found shape (no course matches the query):

{
  "success": false,
  "total_courses": 0,
  "courses": [],
  "error_reasoning": "No course records matched q='<term>' with fq=sectionType_s:course."
}