You'll noticeTier: how much it should matter to you
3Useful: my rating, 1 to 5
2Signal: worth watching, 1 to 5
Web FetchArea: what it touches
ImprovementsKind: in v2.1.307,
ImprovementsSection of the release
What
WebFetch is the tool Claude uses to read a web page. Its results now carry more detail about where the content came from:
Results include a retrieval time (retrievedAt), and the fetch schema gains an optional retrieved_at field.
A result can say the page came from "a stored copy of the page, retrieved from the site on" an earlier date. If that copy is older than a threshold, a note names the retrieval date and warns that the live page may have changed. What supplies the stored copy was not traced.
When a fetch fails at the site itself, WebFetch returns a "Not Fetched" answer instead of page content.
Errors from the CCR proxy, a go-between server that makes requests on Claude Code's behalf, whose source is "target" are now raised as a separate kind of error.
The tool's own description, as captured under this release, is unchanged.
Why
Claude can now tell when web content is old or was never retrieved, and say so to you instead of presenting it as live.
