Source Intelligence
Follow Discord

DisclaimerUnofficial, and not affiliated with Anthropic. Nearly all of this is read straight out of what ships: npm bundles, captured prompts, published docs. Anthropic's own notes go in verbatim, marked as theirs. The rest is my reading, and every entry carries the strings behind it. If one looks wrong, vote it down and say why.

← All of v2.1.234 Home All releases ←olderv2.1.233 v2.1.235newer→
Claude Code v2.1.234

Eval traces identify PDFs and ZIPs by their first bytes

Under the hood
Useful1 Signal1
Eval

Eval traces now recognise PDFs and ZIPs, including Office documents, by their first bytes.

What

A new check reads the leading bytes of a file to spot a PDF signature or a ZIP header, and uses that both to decide the file is binary and to label it, noting that Office documents such as .pptx, .docx and .xlsx are ZIPs.

Evidence

ZIP archive (Office documents such as .pptx/.docx/.xlsx are ZIPs)

Strings lifted out of the shipped bundle, so the claim above can be checked against them.

Related

Other releases about the same thing. Found by shared names or similar wording; neither means one caused the other.

See this entry in the whole of v2.1.234 →