OpenGraph.io

@OpenGraph.io pull the Open Graph metadata for https://example.com/blog/new-feature so we can preview how it looks when shared

Here's the Open Graph metadata for that URL: • Title: "Introducing Our New Dashboard Feature" • Description: "A comprehensive guide to our redesigned analytics dashboard with real-time collaboration tools" • Image: https://example.com/assets/dashboard-preview.png (1200×630) • Type: article • Site Name: Example Blog This metadata controls how the link appears when shared on social platforms like Slack, LinkedIn, or Twitter. The image dimensions meet Open Graph's recommended 1.91:1 aspect ratio.

This example shows the MCP's core metadata extraction capability. Requires a valid API key with sufficient credits. The response reflects what the target site's webmaster configured in their <meta> tags — if those tags are missing or malformed, the output will be sparse or fall back to generic HTML title/description elements.

@OpenGraph.io capture a screenshot of https://competitor-site.com/pricing in desktop viewport so we can compare their tier structure to ours

Screenshot captured successfully. The image shows a three-tier pricing table (Starter at $29/mo, Professional at $99/mo, Enterprise with custom pricing). Each tier lists 5-7 feature bullets. A prominent "Start Free Trial" CTA appears above the table. The page uses a blue-and-white color scheme with the company logo in the top-left corner. Screenshot URL: https://opengraph.io/api/screenshots/abc123def456.png (valid for 30 days) Note: This is a static snapshot taken at the current moment — pricing or features may change after capture.

This demonstrates the screenshot tool's utility for competitive research or visual archiving. The MCP returns a temporary image URL, not the raw bytes, so you'll need to download it if long-term storage is required. Screenshot requests count against your API quota and may take 3-5 seconds to render, especially for JavaScript-heavy pages.

@OpenGraph.io scrape https://docs.example.com/api/webhooks and summarize the authentication requirements mentioned in the page content

I scraped the page and extracted both the raw HTML and metadata. Here's what the documentation says about webhook authentication: The page requires all webhook endpoints to: 1. Validate a signature header (X-Webhook-Signature) using HMAC-SHA256 with your secret key 2. Respond with a 200 status within 5 seconds to confirm receipt 3. Implement replay attack protection by checking the X-Webhook-Timestamp header The page also includes a code example in Python showing how to verify the signature. The Open Graph metadata indicates this is part of a larger API reference (title: "Webhook Authentication | API Docs").

This example pairs the MCP's full-page scraping with the AI's ability to parse and summarize technical content. The scrape tool returns both structured metadata and raw HTML, making it useful when you need more than just social preview tags. Be aware that scraping respects robots.txt and may fail on sites with aggressive bot protection.

When this MCP wins for newsletter and social scheduling

A 3-person content team publishes a weekly newsletter with 8-12 curated links. They draft in Notion, but manually copying titles and descriptions from each URL burns 20 minutes per issue. The Get Open Graph Metadata tool pulls title, description, and hero image in one call—no browser tabs, no copy-paste. This MCP is the right call if your team shares links in any recurring format (newsletters, Slack digests, client reports). The API key setup takes 2 minutes. The threshold: if you're also summarizing or fact-checking the linked content, you'll want a separate web-scraping MCP with full-text extraction. For pure metadata at volume, OpenGraph.io is the fastest path from URL to preview card.

When screenshot and metadata tracking beats manual checks

A 6-person sales team tracks 15 competitor landing pages for pricing and messaging changes. Today they bookmark URLs and eyeball them weekly. The Screenshot Site tool captures a timestamped visual snapshot on demand, and Extract Site Metadata logs the page title and description for diff tracking. This MCP is the right call if your competitive set is under 30 pages and you check them weekly or monthly—not daily. The auth is straightforward (one API key for the workspace). The trade-off: OpenGraph.io doesn't parse pricing tables or run diffs automatically; you're still doing the comparison work in Switchy or a spreadsheet. If you need structured data extraction (pricing, feature lists), a dedicated scraper MCP is the better fit. For visual + metadata snapshots, this is the low-friction option.

When this MCP catches broken external references in docs

A 4-person support team maintains 80 help articles with external links to vendor docs, partner sites, and regulatory pages. Broken links erode trust, but manually checking them quarterly is a 3-hour slog. The Scrape Site tool fetches the raw HTML and metadata in one call—if the request fails or returns a 404, the link is dead. This MCP is the right call if your knowledge base is under 200 external URLs and you audit them monthly or quarterly. The API key lives in Switchy's workspace settings; no per-user config. The threshold: if you need full-text change detection (did the vendor rewrite their API docs?), you'll want a scraper that stores and diffs content. For link-health checks and metadata validation, OpenGraph.io is the simplest tool that works.