broken-link-checker

--- name: broken-link-checker description: Scan documentation for broken links using curl. Use when validating markdown docs, checking for dead URLs, or before releasing documentation. --- # Broken Link Checker Skill Checks for broken or invalid links in documentation and markdown files. ## Usage ```bash # Check current directory node skills/broken-link-checker/index.js # Check specific directory node skills/broken-link-checker/index.js /path/to/docs ``` ## Features - Recursively scans all markdown files in a directory - Extracts both markdown-style links `[text](url)` and raw URLs - Uses curl to check HTTP status codes - Skips anchor links, relative paths, and non-HTTP URLs - Reports file location and line number for each broken link - Configurable timeout and retry logic ## Output ``` === Broken Link Checker === Target: /path/to/project Time: 2026-06-07T09:26:39.852Z Summary: Files scanned: 12 Links found: 47 Links checked: 38 Broken links: 2 Skipped: 9 Broken Links: docs/api.md:15 URL: https://example.com/api/v2/users Status: 404 README.md:42 URL: https://old-docs.example.com/guides Status: ERROR: connect ETIMEDOUT Recommendations: - Found 2 broken link(s) - fix or remove them ``` ## API ### checkBrokenLinks(targetDir, options) Returns check results object: ```javascript { timestamp: '2026-06-07T09:26:39.852Z', target: '/path/to/project', summary: { files_scanned: 12, links_found: 47, links_checked: 38, broken_links: 2, skipped_external: 9 }, files: [...], broken: [...], recommendations: [...] } ``` ## When to Use - Before releasing documentation - During documentation reviews - In CI/CD pipelines for docs - Periodic link health checks - After migrating documentation to new URLs

存档时间线

版本索引永久保留；内容副本只保留最新 1 份，更早版本仅留索引与哈希（存档时间线的证据链不会因此断裂）。

下载存档内容副本