Find archived website versions with AI
Look up captures of a public page, check the actual replay date and keep a defensible record when an index is unavailable.
Find archived website versions with AI when a source has changed and you need to establish what was visible at a particular time. The internet-archive-wayback skill distinguishes an index entry from the page that actually opens. A replay request can redirect to a nearby capture, so the final timestamp matters.
For a historical review of a local festival, begin with the former programme page and a short date range. If the index is returning an outage page, record that rather than concluding the programme was never saved. Another public archive may offer supporting evidence, but it covers a different set of websites.
What it does
- Narrow a capture lookup to an exact URL and period.
- Record the timestamp of the replay that actually opens.
- Distinguish an unavailable index from an absent capture.
- Compare two saved versions of a public notice.
- Use a second archive as a supplement when its coverage fits the source.
Install and use
- Use an HTTP-capable agent and a public URL you are entitled to query. Local Python or shell tools are optional.
- Specify the period and the evidence you need, then request a bounded index lookup.
- Inspect the replay, record its final source and date and note any missing sections.
FAQ
Source
Made by Andrew Lih (fuzheado). Original skill: internet-archive-wayback, published on GitHub under the MIT licence.
View on GitHub