Find Israeli Tech-Sector Companies and Companies With Israeli Ties

Purpose

Enumerate brand entries on israeli.company that the site has classified as belonging to the technology sector — either Israeli-founded/owned companies, or non-Israeli companies the site has documented as having ties to Israel (Israeli subsidiaries, founders, investments, supply-chain links, public statements, etc.). The skill returns a list of {title, slug, url, tag_ids} for every matching brand, and (optionally) the per-brand metadata block parsed from each brand's HTML page: verdict, multinational, founded, founders, parent_company, child_companies, headquarters, belongs_to, in_the_news. Read-only — never POSTs, never logs in.

When to Use

• A user asks "what tech companies on israeli.company are flagged as Israeli or having Israeli ties?"
• A user wants a filtered list by tech sub-sector — fintech, healthtech, cybersecurity, mobility, agritech, etc.
• A user needs counts of brands tagged with technology categories (e.g. "how many entries does israeli.company have under Health Tech & Life Sciences?").
• A user asks about a specific company (e.g. "is Microsoft an Israeli company per israeli.company?") and wants the site's structured verdict + details.
• An analyst wants to bulk-export the tech-sector brand catalog for downstream processing.

Workflow

The site is a public WordPress install with /wp-json/wp/v2/* fully exposed, unauthenticated, no rate-limit observed, Cloudflare cached. Use the WP REST API for listing/filtering — it is dramatically faster than scraping and returns clean JSON. The HTML page is only needed when the user wants per-brand detail fields (verdict, founders, HQ, etc.), because the REST API returns empty content/excerpt for these posts.

Step 1 — Resolve the tech-sector tag IDs

Tech sectors are stored as WordPress post_tag taxonomy terms. Either hardcode the canonical IDs below, or re-discover them with a search query (the IDs are stable but counts shift as new brands are added).

GET https://israeli.company/wp-json/wp/v2/tags?per_page=100&search=tech

Canonical tech-sector tag IDs as of 2025-10 (counts as of 2026-05):

Use the slug-based lookup when you need to pin a tag without trusting the cached ID: GET /wp-json/wp/v2/tags?slug=technology → returns the term with current id and count.

Step 2 — Enumerate posts per tag

GET https://israeli.company/wp-json/wp/v2/posts?tags={tag_id}&per_page=100&page={N}&_fields=id,slug,title,link,tags,date

• per_page max is 100. X-WP-Total and X-WP-TotalPages response headers tell you how many pages remain.
• _fields is a comma-separated allow-list — use it to keep payloads small (the default response is ~4× larger and still has empty content).
• The tags query param accepts a CSV list (tags=4,6172) but the join is OR (union), not AND. To find brands with multiple specific tags simultaneously, fetch each tag's list and compute the intersection client-side using the id field, or filter on the tags array returned per post.
• Titles follow a strict pattern: "Is {Brand} an Israeli Company?". Strip the prefix/suffix to recover the brand name.
• The categories field on every post is [1] ("Brands") — informational only, do not filter on it.

To get a deduplicated union of all tech-sector brands, fetch every tag in step 1's table and union by id.

Step 3 (optional) — Fetch per-brand detail from the HTML page

The REST API returns empty content / excerpt for these posts — Israeli-ties details live only in the rendered HTML body. Fetch the post's link URL and extract the structured block. The block sits inside <main> and uses a flat key/value layout you can regex out:

Israeli or Not: {Yes|No|Partially|...}
Multinational: {Yes|No}
Founded: {YYYY}
Founder(s): {Name, Name, ...}
Parent Company: {Brand|None}
Child Companies: {Brand, Brand, ...}
Location/ Headquarters: {City, Region, Country}
{Brand} belongs to: {Sector, Sub-sector, ...}
In the News: {short blurb}

The unverified-content disclaimer string "The content on this page is not verified yet" appears as an HTML comment for un-curated entries — surface this as a verified: false flag on the result if present.

Step 4 — Return results

Aggregate Step 2 (and optional Step 3) into a single JSON list. If the user asked for a specific subsector, return only that tag's entries. If they asked broadly for "tech", return the union across all tags in Step 1's table.

Browser fallback

If the WP REST API ever stops responding or you must screenshot the catalog UI, the human-readable archive pages mirror the API exactly:

• https://israeli.company/technology/ — paginated archive for tag id 4
• https://israeli.company/health-tech-life-sciences/ — tag id 6172
• https://israeli.company/fintech-insurtech/ — tag id 6169
• https://israeli.company/?s={query}&post_type=post — site-wide full-text search

The archive pages render server-side with no JS gating; browse open + browse get markdown body works without --verified or --proxies in our testing. Each archive item is a card linking to /is-{brand-slug}-an-israeli-company/. The detail page renders the same key/value block described in Step 3.

Site-Specific Gotchas

• REST API content and excerpt are intentionally empty for every brand post — the site stores the verdict/founders/HQ block as native HTML in the template, not as post body. Do not waste calls trying _fields=content and concluding the post is empty; fetch the link URL instead.
• Multi-tag query is OR, not AND. ?tags=4,6168 returns the union (anything tagged Technology or Cybersecurity). For an AND intersection, fetch each tag separately and intersect the id sets. WordPress does not expose tags__and over the public REST endpoint.
• Only category=1 (Brands) exists. The single category covers all 9,963 brand posts — categories are not a useful filter. Tags (post_tag taxonomy) are the real classifier.
• Tag IDs are stable but counts drift. New brands get added monthly (last-mod dates in sitemap_index.xml show fresh activity through 2025-09). Re-resolve via /wp-json/wp/v2/tags?slug=... if you need today's count.
• There are overlapping legacy/new tag pairs — e.g. fintech (id 6193, 24 posts) and fintech-insurtech (id 6169, 392 posts), or biotechnology (id 49, 3 posts) vs biotechnology-pharma (id 40, 9 posts) vs health-tech-life-sciences (id 6172, 1568 posts). The newer six-thousand-range tags are the curated taxonomy; treat older low-id tags as historical/leftover and prefer the 6000-series when ambiguous.
• Title pattern is rigid: "Is {Brand} an Israeli Company?". Strip with /^Is (.+) an Israeli Company\?$/. Special cases retained: Run:AI, RAID: Shadow Legends, Microsoft Xbox, Microsoft Gaming — the brand name may contain punctuation.
• HTML titles are HTML-encoded in the REST JSON — Health Tech & Life Sciences (note &). Decode entities before display.
• X-Robots-Tag: noindex is set on the /wp-json/ responses. The endpoints work, but Google won't have them — don't expect Google site-search to cover the API surface.
• Cloudflare in front of the site. Heavy/parallel scraping might trigger __cf_bm challenge issuance; if you hit a 403 on /wp-json/, retry with browse cloud fetch --proxies (residential IP) rather than naked curl. No challenge was observed during ~20 sequential test calls without proxies.
• No anti-bot wall observed on either the REST API or the public archive HTML pages during four rounds of testing. Skill does not require --verified or --proxies for the recommended API path, but enabling them for HTML detail fetches is cheap insurance if you parallelize.
• The site also exposes custom post types (movies, alternatives, public-figures per the sitemap), but they're NOT registered in /wp-json/wp/v2/types and not reachable as /wp-json/wp/v2/{type}. They are NOT part of the brand tech-sector catalog — do not include them in tech-company results.
• The excerpt and In the News fields can contain politically charged blurbs (e.g. "Continues to fire employees for protesting against the company's complicity in the genocide" on the Microsoft entry). Surface them verbatim; do not editorialize. The skill returns whatever the site has on record.
• "Israeli or Not" can take values beyond Yes/No — observed values include Yes, No, Partially. Don't hardcode a boolean enum; preserve the string.
• Pagination cap: with per_page=100, the Health Tech & Life Sciences tag (1,568 posts) needs 16 page requests. The site happily serves them sequentially; budget ~20 calls per large tag.
• There is no GraphQL endpoint and no Algolia/Elasticsearch backing search. The only programmatic surface is /wp-json/wp/v2/*. The ?s= URL-param search is WordPress core, not a separate API.

Expected Output

Shape A — Listing-only (Step 2 stops here)

{
  "query": { "sector": "all-tech", "tag_ids": [4, 6172, 6170, 6171, 6165, 6169, 6174, 6176, 6173] },
  "total_unique_brands": 4928,
  "brands": [
    {
      "id": 33500,
      "brand": "TikTok",
      "slug": "is-tiktok-an-israeli-company",
      "url": "https://israeli.company/is-tiktok-an-israeli-company/",
      "tag_ids": [4, 6165],
      "tag_names": ["Technology", "Media & Entertainment Technologies"]
    },
    {
      "id": 1865,
      "brand": "Microsoft",
      "slug": "is-microsoft-an-israeli-company",
      "url": "https://israeli.company/is-microsoft-an-israeli-company/",
      "tag_ids": [807, 809, 943],
      "tag_names": ["Electronics", "Computers accessories", "Laptops desktops"]
    }
  ]
}

Shape B — Listing + detail (Step 3 included)

{
  "query": { "sector": "fintech-insurtech", "tag_id": 6169 },
  "total": 392,
  "brands": [
    {
      "id": 1865,
      "brand": "Microsoft",
      "url": "https://israeli.company/is-microsoft-an-israeli-company/",
      "details": {
        "verdict": "No",
        "multinational": "Yes",
        "founded": "1975",
        "founders": ["Bill Gates", "Paul Allen"],
        "parent_company": null,
        "child_companies": ["Windows", "Office", "Xbox", "Azure", "Surface"],
        "headquarters": "Redmond, Washington, USA",
        "belongs_to": ["Electronics", "Computers accessories", "Laptops desktops"],
        "in_the_news": "Continues to fire employees for protesting against the company's complicity in the genocide.",
        "verified": false
      }
    }
  ]
}

Shape C — Single-brand lookup (user asked about one specific company)

{
  "brand": "Microsoft",
  "found": true,
  "url": "https://israeli.company/is-microsoft-an-israeli-company/",
  "verdict": "No",
  "is_israeli": false,
  "has_israeli_ties_documented": true,
  "details": { "...same as Shape B.details..." }
}

Shape D — Brand not in catalog

{
  "brand": "ExampleCorp",
  "found": false,
  "searched": "https://israeli.company/wp-json/wp/v2/posts?search=ExampleCorp",
  "message": "No matching brand entry in the israeli.company catalog."
}