⌘ SEO Evaluation API for Cursor IDE

Give your Cursor AI agent the ability to test SEO quality of any page it edits.

Cursor is the AI-native code editor. By adding a custom rule, you can teach Cursor’s AI to automatically evaluate SEO quality whenever it generates or modifies web pages. The agent calls SEODiff’s agentic evaluation API and interprets the results inline.

System Prompt / Configuration

Copy this prompt and paste it into your IDE’s AI configuration:

# SEO Quality Rule for Cursor
# Add this to your .cursor/rules file or Cursor Settings > Rules for AI

Whenever you create or edit an HTML page, a Next.js page component, or a
template that renders HTML:

1. After making the change, call the SEODiff API to evaluate the page:
   POST https://seodiff.io/api/v1/agent/evaluate
   Headers: Authorization: Bearer $SEODIFF_API_KEY
   Body: {
     "urls": ["<the URL of the page you edited>"],
     "assertions": [
       {"rule": "has_h1"},
       {"rule": "has_meta_description"},
       {"rule": "has_schema"},
       {"rule": "no_placeholders"},
       {"rule": "min_word_count", "value": 200},
       {"rule": "max_token_bloat", "value": 8.0}
     ],
     "wait": true
   }

2. If any assertion fails, fix the issue before considering the task complete.
3. Report the ACRI score and any failing assertions to the user.

Setup

# Set your API key in Cursor's environment
# Cursor Settings > Features > Terminal > Environment Variables
SEODIFF_API_KEY=your_api_key_here

Example Interaction

User: "Create a landing page for our new product, SuperWidget."

Cursor AI:
1. Creates the page with H1, meta description, schema markup
2. Calls SEODiff API: POST /api/v1/agent/evaluate
3. Result: 5/6 assertions pass, "has_schema" failed (missing Product schema)
4. Adds JSON-LD Product schema
5. Re-evaluates: 6/6 pass, ACRI score: 78
6. Reports: "Page created with ACRI score 78. All SEO assertions pass."