Skip to main content
Glama
lakshyamehta03

mcp-server-wayback

by lakshyamehta03
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Python
Remote

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
WAYBACK_MCP_IA_ACCESS_KEYNoYour Internet Archive S3 access key for higher rate limits.
WAYBACK_MCP_IA_SECRET_KEYNoYour Internet Archive S3 secret key for higher rate limits.

Instructions

Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.

This server publishes no instructions, or was last inspected before Glama recorded them.

Capabilities

Features and capabilities supported by this server

Protocol revision2025-11-25

CapabilityDetails
tools
{
  "listChanged": false
}
prompts
{
  "listChanged": false
}
resources
{
  "subscribe": false,
  "listChanged": false
}
experimental
{}

Tools

Functions exposed to the LLM to take actions

NameDescription
check_availabilityA

Check if a URL has been archived by the Wayback Machine and return the closest snapshot.

lookup_snapshotsA

Return CDX snapshots for a URL, with optional date range and status-code filter.

The Wayback Machine often crawls the same URL many times per day; raw CDX results would return one row per crawl. collapse is a server-side de-duplication: adjacent rows that share the same value in the chosen field get folded into a single representative row.

By default we collapse on the first 8 digits of the timestamp ("timestamp:8"), which is the YYYYMMDD prefix — i.e. one row per day. This is almost always what you want for "show me snapshots of this URL"; otherwise the default limit of 50 gets eaten by 50 captures from a single hour and you see nothing about the URL's history.

Override collapse when you need different granularity:

  • "digest" — collapse on content hash, so you only see captures where the page actually changed

  • "timestamp:10" — one row per hour (first 10 digits of timestamp)

  • "" (empty string) — disable collapsing entirely; return every capture

  • any other CDX collapse spec is passed through verbatim

latest=True uses CDX's fastLatest path to return the N most recent captures cheaply (much faster than a full scan over the index). Cannot be combined with from_date/to_date.

search_archiveA

Search Internet Archive collections (uploaded books, audio, video, software items) using Lucene query syntax.

This is NOT a search over the Wayback Machine web crawl. It only returns items that someone has uploaded to archive.org as a discrete media item.

Do NOT use this for:

  • Current news, journalism, or recent events

  • Government circulars, press releases, or official web pages

  • Wikipedia articles or any live web content

  • "What was on this website" / "what URLs are archived" — use search_domain or lookup_snapshots for those.

Good uses: historical books, lecture recordings, archived films, software releases, podcast episodes, scanned magazines. Use Lucene fields when possible (e.g. subject:"civil war", creator:"NASA", collection:librivoxaudio).

search_domainC

Find archived URLs under a domain or path prefix. Auto-detects matchType from input.

get_snapshot_contentA

Fetch and extract text content from an archived web page. Returns extracted text, word count, and extraction metadata.

get_item_metadataC

Fetch rich structured metadata for any Internet Archive item by its identifier.

Prompts

Interactive templates invoked by user choice

NameDescription
research_topicResearch a topic across Internet Archive collections and synthesise an overview. Takes a single free-text `topic` argument. Time ranges and sample sizes are interpreted from the topic phrasing by the model (e.g. "the moon landing in the 1960s" → year range 1960–1969).
track_site_changesNarrate how an archived web page changed over time using sampled snapshots. Takes a single free-text `url` argument. Date ranges are interpreted from the surrounding context by the model when present.
audit_link_rotAudit a list of URLs for link rot, surfacing archived alternatives.
setup_authenticationWalk the user through configuring Internet Archive API keys for higher rate limits.

Resources

Contextual data attached and managed by the client

NameDescription

No resources

Deploy Server
One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

TDQS

A3.7/5.0

Scored across 6 tools

Disambiguation5/5

Each tool has a distinct purpose: checking availability, fetching metadata, extracting text, looking up CDX snapshots, searching collections, and searching domains. No overlapping functionality.

Naming Consistency5/5

All tool names follow a consistent verb_noun snake_case pattern (e.g., check_availability, get_item_metadata, lookup_snapshots), making them predictable and easy to distinguish.

Tool Count5/5

Six tools is appropriate for a Wayback Machine / Internet Archive server, covering the main operations without being overwhelming or too sparse.

Completeness4/5

The set covers core read operations (check, get metadata, get content, lookup snapshots, search archive items, search domains). Missing a 'save page' tool and raw HTML retrieval, but covers expected use cases well.

Maintenance

ActivityInactive
ResponsivenessResponsive

Resources