Jerusalem Municipality — Council Committee Meetings

Access archived committee sittings (ישיבות ועדה) of Jerusalem City Council: list meetings, open a specific meeting's agenda, and download the protocol PDF.

URLs

• Archive / landing page — list of committees and their meetings: https://www.jerusalem.muni.il/he/city/council/committeesmeetings/
• Single meeting detail page — query-string driven: https://www.jerusalem.muni.il/he/city/council/CommitteesMeetings/CommitteeMeeting?term=<TERM>&id=<GUID>
  • term — council term number (e.g. 17).
  • id — meeting GUID (e.g. 85cf59bc-f6d0-f011-bbd3-7ced8d49cdf0).

Access notes

• Akamai edge protection: plain curl / fetch returns HTTP 403 ("Access Denied" from errors.edgesuite.net) even with a browser User-Agent. Use a real browser — Playwright (headless is usually fine; fall back to headed if Akamai challenges).
• Geo: not confirmed as strictly IL-geo-restricted — the 403 is bot-detection, not geo. Treat as tier 2 (remote headless browser) preferred, tier 3 (local Playwright) as fallback.
• Language: Hebrew (RTL). No login required for public protocols.

What to extract from a meeting detail page

From CommitteeMeeting?term=<N>&id=<GUID>:

The agenda outline on the page is the cheap summary; the פרוטוקול PDF is the authoritative record.

Procedure

1. Resolve target meeting. User supplies either:
   • A full CommitteeMeeting?term=…&id=… URL, or
   • A description ("latest finance committee", "planning committee from March 2026") — in which case start at the archive landing page and pick the matching row.
2. Open the detail page with Playwright. Default to headless Chromium. If the page renders "Access Denied" or a challenge, retry headed.
3. Extract the agenda — grab the section headings and any bullet items under סדר יום / נושאים. Preserve Hebrew verbatim.
4. Find the protocol link. Look for an anchor whose visible text contains פרוטוקול (or an href ending in .pdf near that label). Resolve to an absolute URL.
5. Download the PDF. Use the same browser context (cookies/session) — page.request.get(pdf_url) in Playwright, or click-and-capture-download. Save to the user's requested path, defaulting to: ~/jerusalem-council-meetings/<term>-<short-id>-<YYYY-MM-DD>.pdf
6. Report back — committee name, date, meeting number, agenda outline, and the saved PDF path.

Playwright sketch

from playwright.sync_api import sync_playwright
from pathlib import Path

URL = "https://www.jerusalem.muni.il/he/city/council/CommitteesMeetings/CommitteeMeeting?term=17&id=85cf59bc-f6d0-f011-bbd3-7ced8d49cdf0"

with sync_playwright() as p:
    browser = p.chromium.launch(headless=True)
    ctx = browser.new_context(locale="he-IL")
    page = ctx.new_page()
    page.goto(URL, wait_until="domcontentloaded")

    # Agenda
    agenda = page.locator("main").inner_text()

    # Protocol PDF — anchor with text containing פרוטוקול, or href ending .pdf
    pdf_link = page.locator("a:has-text('פרוטוקול')").first
    pdf_url = pdf_link.get_attribute("href")
    if pdf_url and pdf_url.startswith("/"):
        pdf_url = "https://www.jerusalem.muni.il" + pdf_url

    # Download through the browser context so cookies/Akamai tokens are reused
    resp = ctx.request.get(pdf_url)
    out = Path.home() / "jerusalem-council-meetings" / "17-85cf59bc.pdf"
    out.parent.mkdir(parents=True, exist_ok=True)
    out.write_bytes(resp.body())

    browser.close()

Selectors above are best-effort from the described structure — verify on first run and adjust if the markup differs (especially the פרוטוקול anchor and any cookie/consent banner that may intercept clicks).

Output format

Committee: <HE name>  (<English gloss if obvious>)
Date:      <YYYY-MM-DD>
Term:      <N>    Meeting #: <M>
Agenda:
  1. <section heading>
     - <bullet>
     - <bullet>
  2. ...
Protocol PDF saved to: <absolute path>   (<size>, <n> pages)

Out of scope

• Solving Akamai CAPTCHA / challenges (if one appears, surface it to the user in a headed browser).
• Parsing the PDF contents — this skill only fetches the file. Combine with a separate PDF-reading step for content extraction.
• Non-committee council outputs (full council votes, budget documents) — those live on other sections of the site.