find-latest-news

name

find-latest-news

description

Find the latest news, press coverage, and social signals for each sourced YC company. Uses Exa news search to surface recent TechCrunch articles, launch announcements, product launches, hiring news, and founder interviews. Triggers on: "find news", "get latest news", "find press coverage", "find articles", "step 4", "next step", "continue to find news", "get news about companies", "check news", "find recent coverage".

Find Latest News

Pipeline: source-ycombinator → enrich-founders → scrape-website → find-latest-news → find-latest-fundraise → export-csv

You are step 4 of 5. Your job is to find the latest press coverage, product announcements, and signals of momentum for each YC company.

Tools

• Exa MCP (web_search_exa) — primary: find recent news, articles, and press coverage
• Exa MCP (crawling_exa) — fetch full article content when summary is insufficient
• WebSearch / WebFetch — fallback for any source Exa doesn't cover

Input

Read data/raw/yc_companies.json. Use company_name and website as search keys.

Step 1: Search for recent news per company

For each company, run at least 2 Exa searches targeting different news angles:

Query 1 — General coverage:

web_search_exa("[company_name] startup", date_range: last 12 months)

Query 2 — Launch and product news:

web_search_exa("[company_name] launch OR product OR raises", date_range: last 12 months)

Query 3 — Founder press (optional, for high-signal companies):

web_search_exa("[founder_name] [company_name]", date_range: last 24 months)

Prioritise sources: TechCrunch, Forbes, Business Insider, VentureBeat, The Information, Bloomberg, Reuters, Product Hunt, Hacker News.

Step 2: Filter and score articles

For each article found, extract:

• title — article headline
• url — article URL
• source — publication name (e.g. TechCrunch)
• published_date — date published (ISO format)
• summary — 1–2 sentence summary of what the article says
• signal_type — classify the news type:
  • funding — fundraising announcement (skip here — handled by find-latest-fundraise)
  • product_launch — new product or feature launch
  • partnership — partnership or enterprise deal announced
  • hiring — significant hiring or team expansion
  • award — award, recognition, or ranking
  • founder_interview — founder profile or interview
  • customer_win — notable customer or case study
  • general — general coverage

Keep up to 5 most recent and relevant articles per company. Exclude pure funding announcements (those go in find-latest-fundraise).

If no articles are found, set news: [] — do not fabricate coverage.

Step 3: Derive momentum signals

Based on the news found, assign a news_momentum score:

Criteria	Score
No coverage	0
1–2 articles, > 6 months old	1
1–2 articles, < 6 months old	2
3+ articles in last 6 months	3
TechCrunch / major press in last 3 months	4
Multiple major press + product launch in last 3 months	5

Also flag:

• product_hunt_featured: true/false — if Product Hunt coverage found
• hacker_news_featured: true/false — if HN Show/Ask coverage found

Output

Write an array to data/normalized/company_news.json conforming to schemas/company_news.schema.json.

One record per company, including company_id and company_name as join keys.

Review checkpoint

After writing data/normalized/company_news.json, present a summary:

Found news coverage for [N] / [total] companies:

  No coverage: [N] companies
  Low momentum (1–2): [N] companies
  Strong momentum (3+): [N] companies
  Product Hunt featured: [N]
  Hacker News featured: [N]

Top stories by recency:
| Company | Source | Date | Signal |
|---------|--------|------|--------|
| ... | TechCrunch | 2025-04-12 | product_launch |
...

Does this look right? You can:
  - Confirm to proceed to step 5 (find latest fundraise)
  - Add a news article manually: "add article for [company name]: [url]"
  - Re-search a company: "re-search news for [company name]"

Do not proceed to find-latest-fundraise until the VC confirms.