AgentSkillsCN

wayback-screenshot

从 Wayback Machine 中获取网页截图。当用户想要查看网页的外观、获取视觉快照、查找归档截图，或回顾网页的历史面貌时，可使用此功能。

SKILL.md
--- frontmatter
name: wayback-screenshot
description: Retrieve screenshots from the Wayback Machine. Use when the user wants to see how a webpage looked, get a visual snapshot, find archived screenshots, or view historical page appearance.

Retrieve Wayback Machine Screenshots

Access existing screenshots stored by the Wayback Machine.

Usage

bash
npx tsx scripts/screenshot.ts <url> [options]

Arguments

ArgumentRequiredDescription
urlYesThe URL to find screenshots for

Options

OptionDescription
--timestamp=DATEGet screenshot from specific capture (YYYYMMDDhhmmss)
--listList available screenshots for URL
--download=PATHDownload screenshot to file
--no-cacheBypass cache and fetch fresh data from API

Output

code
Screenshots for: https://example.com/

January 15, 2024 12:34 (3 days ago)
  https://web.archive.org/web/20240115123456im_/https://example.com/

December 1, 2023 08:00 (46 days ago)
  https://web.archive.org/web/20231201080000im_/https://example.com/

Total: 2 screenshot(s)

Script Execution (Preferred)

bash
npx tsx scripts/screenshot.ts <url> [options]

Options:

  • •--timestamp=DATE - Get screenshot from specific capture (YYYYMMDDhhmmss)
  • •--list - List available screenshots for URL
  • •--download=PATH - Download screenshot to file
  • •--no-cache - Bypass cache and fetch fresh data from API

Run from the wayback plugin directory: ~/.claude/plugins/cache/wayback/

Screenshot URL Pattern

code
https://web.archive.org/screenshot/{URL}

Direct Access

Get the most recent screenshot:

code
https://web.archive.org/screenshot/https://example.com/

With Timestamp

Get screenshot from a specific capture:

code
https://web.archive.org/web/{timestamp}im_/https://example.com/

The im_ modifier returns the screenshot image for that timestamp.

List Available Screenshots

Use CDX API with wildcard to find all screenshots:

code
https://web.archive.org/cdx/search/cdx?url=web.archive.org/screenshot/https://example.com/*&output=json

Or browse visually:

code
https://web.archive.org/web/*/https://web.archive.org/screenshot/https://example.com/*

Via Wayback Toolbar

When viewing any archived page:

  1. •Look for the camera icon (📷) in the top-right of the Wayback toolbar
  2. •Click to view available screenshots for that capture

From SPN2 Response

When submitting with capture_screenshot=1, the response includes:

json
{
  "status": "success",
  "screenshot": "https://web.archive.org/web/20240115123456im_/https://example.com/"
}
code

## Caveats

- **Not all captures have screenshots** - depends on whether `capture_screenshot=1` was used during archiving
- **Undocumented feature** - may be unreliable or change without notice
- **Indexing delays** - newly captured screenshots may not appear immediately
- **Coverage varies** - older archives typically don't have screenshots

## Caching

Availability API responses are cached for 24 hours using the OS temporary directory (`os.tmpdir()`). Cache keys are generated from the URL using SHA-256 hashing. Cached responses expire automatically and are deleted on access.

Use `wayback-cache` to manage cached data:
```bash
npx tsx scripts/cache.ts clear    # Clear all cache
npx tsx scripts/cache.ts status   # Show cache status

See wayback-cache skill for complete cache management documentation.

Related

  • •Use wayback-submit --capture-screenshot to create a new screenshot
  • •Use wayback-check to verify if a URL is archived
  • •Use wayback-list to see all captures (not just those with screenshots)