Reference

Data source: Wayback Machine

MCP MCP Protocol

Endpoint https://context.gnist.ai/mcp/wayback/

nobloat — strip report_feedback tool URL-based auth (for clients without header support)

claude_desktop_config.json

{
  "mcpServers": {
    "gnist-wayback": {
      "url": "https://context.gnist.ai/mcp/wayback/",
      "headers": {
        "Gnist-API-Key": "YOUR_API_KEY"
      }
    }
  }
}

Shell

curl -X POST "https://context.gnist.ai/mcp/wayback/" \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -H "Gnist-API-Key: YOUR_API_KEY" \
  -d '{"jsonrpc": "2.0", "method": "tools/call", "id": 1, "params": {"name": "check_availability", "arguments": {"url": "example"}}}'

API REST API

API Root https://context.gnist.ai/rest/wayback/

5 endpoints available:

GET /rest/wayback/check_availability
GET /rest/wayback/search_snapshots
GET /rest/wayback/get_capture_calendar
GET /rest/wayback/get_capture_stats
GET /rest/wayback/report_feedback

Shell

curl "https://context.gnist.ai/rest/wayback/check_availability?url=example" \
  -H "Gnist-API-Key: YOUR_API_KEY"

Python

import httpx

resp = httpx.get(
    "https://context.gnist.ai/rest/wayback/check_availability",
    headers={"Gnist-API-Key": "YOUR_API_KEY"},
    params={
        "url": "example"
},
)
print(resp.json())

OpenAPI Spec · Full API Docs

CLI Command Line

Command gc wayback

CLI documentation will be available when the GnistAI SDK is published.

Install

pip install gnistai

Available Tools (5)

check_availability

Check if a URL has been archived in the Wayback Machine and find the closest snapshot.

Parameter	Type	Required	Description
`url`	`string`	required	The URL to look up (e.g. "https://example.com/page").
`timestamp`	`any`	optional	Optional target date/time (YYYY-MM-DD or YYYYMMDD). When provided, the Wayback Machine returns the snapshot closest to that date. Omit to get the most recent snapshot.

search_snapshots

Search the Wayback Machine CDX index for archived snapshots of a URL.

Parameter	Type	Required	Description
`url`	`string`	required	The URL to search (e.g. "https://example.com/page"). Use wildcard prefix matching with "example.com/*" to capture all pages on a domain.
`date_from`	`any`	optional	Only include captures on or after this date (YYYY-MM-DD). Optional.
`date_to`	`any`	optional	Only include captures on or before this date (YYYY-MM-DD). Optional.
`limit`	`integer`	optional	Maximum snapshots to return (1-100, default 20). (default: 20)
`status_code`	`any`	optional	Filter by HTTP status at capture time (e.g. "200", "404"). Optional.
`mimetype`	`any`	optional	Filter by MIME type (e.g. "text/html", "application/pdf"). Optional.

get_capture_calendar

Get monthly capture counts for a URL in a given year from the Wayback Machine.

Parameter	Type	Required	Description
`url`	`string`	required	The URL to get capture calendar for (e.g. "https://example.com").
`year`	`integer`	required	Year to retrieve calendar for (e.g. 2023).