Prazo Certo — Check for a New Blog Post

Purpose

Answers the question "Tem um blogpost novo no site do Mário?" — is there a new blog post on Prazo Certo (prazocerto.me), Mário Lúcio's site? This is a read-only skill that returns the newest post's title, publish date, author, and URL, plus a simple has_new_post recency judgement. The fast, reliable path is the site's open WordPress REST API, which returns posts already sorted newest-first as JSON — no scraping, no browser, no anti-bot friction.

When to Use

A user asks whether Mário / Prazo Certo published anything new, or wants "the latest post."
You need the newest post's title / date / link to summarize, share, or monitor.
You want to poll for new content on a schedule (compare the newest id or date against a stored value).
You need the last N posts (change per_page), not just the single newest one.

Workflow

The recommended method is the WordPress REST API. prazocerto.me is a WordPress site (LiteSpeed + Cloudflare) that exposes the standard wp-json API publicly, discoverable via the Link: <https://prazocerto.me/wp-json/>; rel="https://api.w.org/" header on any page.

GET the newest post(s) as JSON. Request only the fields you need with _fields:
```
GET https://prazocerto.me/wp-json/wp/v2/posts?per_page=1&orderby=date&_fields=id,date,modified,slug,link,title
```
- The response is a JSON array already sorted newest-first (orderby=date is the default; kept explicit for clarity). Element [0] is the newest post.
- Increase per_page (max 100) to pull a list; the X-WP-Total / X-WP-TotalPages response headers tell you the full post count (~31 posts / 11 pages of 3 at time of authoring).
- This endpoint returns HTTP 200 from datacenter IPs — it does not require residential proxies or stealth. (browse cloud fetch <url> works with or without --proxies.)
Read the fields from element [0]:
- title.rendered → post title (HTML-entity encoded; decode if displaying).
- date → publish datetime, site-local (America/Sao_Paulo), e.g. 2026-07-25T21:17:27. Take the date part for published_date.
- link → canonical post URL (note: posts live at the site root, e.g. /projetos-praticos/, not under /blog/).
- modified → last-edited datetime (useful to distinguish "new post" from "edited old post").
Judge recency. Compute days_since_published = today − date. Treat has_new_post: true when the newest post is within ~30 days. (Adjust the window to the caller's definition of "new"; document the assumption.)
Author. All posts are authored by Mário Lúcio (the site owner). If you need it programmatically, add author to _fields and resolve via /wp-json/wp/v2/users/<id>, or read the Por Mário Lúcio byline on the post page.

Assumption made when the caller doesn't specify: "new" = published within the last 30 days. State this in your answer.

Browser fallback

Only needed if the REST API is ever disabled. The rendered pages sit behind Cloudflare, so a plain headless hit may be challenged — use a stealth session (--verified --proxies).

browse open https://prazocerto.me/blog/ --remote (verified + proxies session).
browse wait load, then dismiss the "Gerenciar Consentimento de Cookies" modal (click "Aceitar cookies") — it overlays content on first load.
The blog index lists posts newest-first; the first card's title + DD/MM/YYYY byline is the newest post. Open it for the full byline (Mário Lúcio / 25/07/2026 / N comentários).
Do not log in, comment, or submit the contact form. Read-only.

Site-Specific Gotchas

Open WP REST API is the whole game. https://prazocerto.me/wp-json/wp/v2/posts is public, unauthenticated, and returns clean JSON newest-first. Prefer it over scraping the HTML in every case.
Posts are NOT under /blog/. The blog index is /blog/ (paginated: /blog/page/2/ …), but individual posts live at the site root, e.g. https://prazocerto.me/projetos-praticos/. Use the link field from the API rather than constructing URLs.
orderby=date ≠ orderby=modified. Default ordering is by publish date. An old post that was recently edited will still sort as old — check modified if "recently touched" matters more than "newly published."
title.rendered is HTML-entity encoded (e.g. –, accented chars as UTF-8). Decode before display.
Dates are site-local (America/São Paulo) with no timezone suffix. Don't assume UTC when computing days_since_published.
Cloudflare only bites the rendered HTML, not the API. The API returned 200 from a bare datacenter fetch across multiple runs. The /blog/ and post pages are Cloudflare-fronted; drive them with --verified --proxies if you must, and expect a cookie-consent modal.
robots.txt explicitly welcomes AI crawlers (GPTBot, ClaudeBot, PerplexityBot, etc. all Allow: /) and only blocks /admin/, /wp-admin/, /private/. Reading posts is fully sanctioned.
sitemap.xml lists only static pages (/sobre, /portfolio, /blog/, /contato/, /mentorias, …) — not individual posts, and its lastmod values are stale (2025-01-01). Do not use the sitemap to find the newest post; use the API.
Recency window is a judgement call. The API gives you facts (title/date); "new" is subjective. At time of authoring, the newest post was 2026-07-25 (7 days before the 2026-08-01 check) → considered new.

Expected Output

New post found (recent) — the normal case:

{
  "success": true,
  "has_new_post": true,
  "latest_post": {
    "title": "A teoria é incrível, mas aplique isso em projetos práticos",
    "author": "Mário Lúcio",
    "published_date": "2026-07-25",
    "modified_date": "2026-07-25",
    "url": "https://prazocerto.me/projetos-praticos/",
    "days_since_published": 7
  },
  "total_posts": 31,
  "checked_at": "2026-08-01",
  "error_reasoning": null
}

No recent post (newest exists but is older than the recency window):

{
  "success": true,
  "has_new_post": false,
  "latest_post": {
    "title": "Nem todas as experiências são individuais",
    "author": "Mário Lúcio",
    "published_date": "2026-06-09",
    "url": "https://prazocerto.me/nem-todas-as-experiencias-sao-individuais/",
    "days_since_published": 53
  },
  "total_posts": 31,
  "checked_at": "2026-08-01",
  "error_reasoning": null
}

Failure (API unreachable and browser fallback blocked):

{
  "success": false,
  "has_new_post": null,
  "latest_post": null,
  "checked_at": "2026-08-01",
  "error_reasoning": "wp-json returned 403 and /blog/ hit a Cloudflare challenge that did not resolve"
}

pzct-blog