Verify all external links in the specified markdown file: $ARGUMENTS
If no file path is provided, ask the user which file to verify.
Process
Step 1: Extract Links and Context
Read the file and extract all markdown links [text](url) along with their surrounding context.
For each link, capture:
- •The URL
- •The anchor text
- •The surrounding sentence or bullet point describing what the link is about
Step 2: Verify Each Link
For each link:
- •
Check HTTP Status using curl:
bashcurl -s -o /dev/null -w "%{http_code}" "<URL>"- •200: Proceed to content verification
- •301/302: Follow redirect, verify final destination
- •404: Mark as broken
- •Other: Mark as problematic
- •
Verify Content Matches Context - For links returning 200:
- •Use WebFetch to analyze the page content
- •Compare against what the notes claim about it
- •Flag mismatches where page content doesn't support the citation
Step 3: Report Results
Provide a summary:
Verified - Links where HTTP status is good AND content matches the citation
Broken - Links returning errors, with suggested corrections if possible
Content Mismatches - Links where page exists but content doesn't match claims
Redirects - Links that redirect, with final destinations
Example Output
code
## Link Verification Report for notes/2026/01/22/notes.md ### Verified (4 links) - [Granola](https://www.granola.ai/) - Content confirms: meeting transcription tool - [Prodigy](https://prodi.gy/) - Content confirms: annotation tool for ML ### Broken (1 link) - [Apple Vision](https://developer.apple.com/.../poses-in-an-image) - Status: 404 - Suggested: .../poses-in-images (plural) ### Content Mismatches (0 links) None found.